diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13167 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1875,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016,
+      "grad_norm": 2.6508815428813,
+      "learning_rate": 5.319148936170213e-08,
+      "loss": 1.3588,
+      "step": 1
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 2.5457892121592454,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 1.323,
+      "step": 2
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 2.860158400454203,
+      "learning_rate": 1.5957446808510638e-07,
+      "loss": 1.4178,
+      "step": 3
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 2.5661170757514604,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 1.3591,
+      "step": 4
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 2.4777548048018847,
+      "learning_rate": 2.6595744680851066e-07,
+      "loss": 1.3441,
+      "step": 5
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 2.5083846992662666,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 1.4051,
+      "step": 6
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 2.4445323039579585,
+      "learning_rate": 3.723404255319149e-07,
+      "loss": 1.3674,
+      "step": 7
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 2.481088215788748,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.3869,
+      "step": 8
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 2.343270980925809,
+      "learning_rate": 4.787234042553192e-07,
+      "loss": 1.4282,
+      "step": 9
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 2.4661217169738734,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 1.3501,
+      "step": 10
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 2.1984860413478104,
+      "learning_rate": 5.851063829787235e-07,
+      "loss": 1.3827,
+      "step": 11
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 2.4853285507338865,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 1.3667,
+      "step": 12
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 2.412927494212378,
+      "learning_rate": 6.914893617021278e-07,
+      "loss": 1.4221,
+      "step": 13
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 2.249266531013893,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 1.369,
+      "step": 14
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 2.1707507731817586,
+      "learning_rate": 7.97872340425532e-07,
+      "loss": 1.346,
+      "step": 15
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 2.203769482457407,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 1.3548,
+      "step": 16
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 2.5343584450361925,
+      "learning_rate": 9.042553191489363e-07,
+      "loss": 1.4092,
+      "step": 17
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 2.0131524369253215,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 1.3371,
+      "step": 18
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 2.186260933309721,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 1.3581,
+      "step": 19
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 1.631781496512265,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 1.3206,
+      "step": 20
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 1.8778425497539284,
+      "learning_rate": 1.1170212765957447e-06,
+      "loss": 1.3465,
+      "step": 21
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 1.8188788465372485,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 1.3834,
+      "step": 22
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 1.8846208068064163,
+      "learning_rate": 1.223404255319149e-06,
+      "loss": 1.4014,
+      "step": 23
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 1.680939894693523,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.3414,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9958976210384602,
+      "learning_rate": 1.3297872340425533e-06,
+      "loss": 1.4561,
+      "step": 25
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 1.746093510878624,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 1.3571,
+      "step": 26
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 1.7118623585900545,
+      "learning_rate": 1.4361702127659578e-06,
+      "loss": 1.35,
+      "step": 27
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 1.6447275181524579,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 1.3052,
+      "step": 28
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 1.739335007475933,
+      "learning_rate": 1.5425531914893618e-06,
+      "loss": 1.3093,
+      "step": 29
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.628647528466785,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 1.2384,
+      "step": 30
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 2.1276333711019557,
+      "learning_rate": 1.648936170212766e-06,
+      "loss": 1.2914,
+      "step": 31
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 1.7010139668434727,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.2839,
+      "step": 32
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 1.944396399002483,
+      "learning_rate": 1.7553191489361704e-06,
+      "loss": 1.3749,
+      "step": 33
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 1.6831405536910045,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 1.3172,
+      "step": 34
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 1.6835872131936704,
+      "learning_rate": 1.8617021276595745e-06,
+      "loss": 1.3987,
+      "step": 35
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.433952564110348,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 1.3312,
+      "step": 36
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 1.4803732161405265,
+      "learning_rate": 1.968085106382979e-06,
+      "loss": 1.3506,
+      "step": 37
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 1.4217221535717617,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 1.2903,
+      "step": 38
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 1.3982889633462,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 1.2889,
+      "step": 39
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 1.5713570551010418,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 1.3404,
+      "step": 40
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 1.5215024708131804,
+      "learning_rate": 2.1808510638297876e-06,
+      "loss": 1.3035,
+      "step": 41
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 1.6955442513159007,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 1.3802,
+      "step": 42
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 1.4763797501186007,
+      "learning_rate": 2.2872340425531916e-06,
+      "loss": 1.3596,
+      "step": 43
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 1.5181782514073643,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 1.2217,
+      "step": 44
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 1.4046404658793936,
+      "learning_rate": 2.393617021276596e-06,
+      "loss": 1.2767,
+      "step": 45
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 1.462947538022516,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 1.302,
+      "step": 46
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 1.2892961442963045,
+      "learning_rate": 2.5e-06,
+      "loss": 1.2487,
+      "step": 47
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 1.3226533611568942,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 1.1913,
+      "step": 48
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 1.3476093040944925,
+      "learning_rate": 2.6063829787234047e-06,
+      "loss": 1.205,
+      "step": 49
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3221965193129026,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 1.3028,
+      "step": 50
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 1.3958129456251924,
+      "learning_rate": 2.7127659574468084e-06,
+      "loss": 1.2695,
+      "step": 51
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 1.4044669269799819,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 1.3055,
+      "step": 52
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 1.3368096346491132,
+      "learning_rate": 2.819148936170213e-06,
+      "loss": 1.2552,
+      "step": 53
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 1.4033198863641372,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 1.3589,
+      "step": 54
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 1.3297672542955576,
+      "learning_rate": 2.9255319148936174e-06,
+      "loss": 1.2533,
+      "step": 55
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 1.3600581237455072,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.2865,
+      "step": 56
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 1.317119263523914,
+      "learning_rate": 3.031914893617022e-06,
+      "loss": 1.2839,
+      "step": 57
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 1.349993584985242,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 1.229,
+      "step": 58
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 1.3011137840566604,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 1.2645,
+      "step": 59
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 1.3201241706938067,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 1.3262,
+      "step": 60
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 1.2697377429716319,
+      "learning_rate": 3.24468085106383e-06,
+      "loss": 1.2183,
+      "step": 61
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 1.2696547481303002,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 1.2516,
+      "step": 62
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 1.4288617262875076,
+      "learning_rate": 3.3510638297872345e-06,
+      "loss": 1.3512,
+      "step": 63
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 1.3772737398632353,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.2807,
+      "step": 64
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 1.2658916424347746,
+      "learning_rate": 3.457446808510639e-06,
+      "loss": 1.2433,
+      "step": 65
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 1.334567599068373,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 1.3031,
+      "step": 66
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 1.276484222880141,
+      "learning_rate": 3.5638297872340426e-06,
+      "loss": 1.2172,
+      "step": 67
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 1.3914548666219049,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 1.2694,
+      "step": 68
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 1.2449140907541763,
+      "learning_rate": 3.670212765957447e-06,
+      "loss": 1.3043,
+      "step": 69
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 1.418974878409878,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 1.2603,
+      "step": 70
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 1.3334985658777239,
+      "learning_rate": 3.7765957446808516e-06,
+      "loss": 1.2131,
+      "step": 71
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 1.2696952693157317,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.205,
+      "step": 72
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 1.4830510678810025,
+      "learning_rate": 3.882978723404256e-06,
+      "loss": 1.2363,
+      "step": 73
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 1.2832900079781504,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 1.2799,
+      "step": 74
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3141641695371142,
+      "learning_rate": 3.98936170212766e-06,
+      "loss": 1.2316,
+      "step": 75
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 1.3384629144912241,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 1.2497,
+      "step": 76
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 1.2305016222469252,
+      "learning_rate": 4.095744680851064e-06,
+      "loss": 1.2229,
+      "step": 77
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 1.3122289104554492,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 1.28,
+      "step": 78
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 1.3216635013853866,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 1.2372,
+      "step": 79
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 1.2173858137755214,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.2862,
+      "step": 80
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 1.2727116908067995,
+      "learning_rate": 4.308510638297873e-06,
+      "loss": 1.1822,
+      "step": 81
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 1.2826192505493774,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 1.2519,
+      "step": 82
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 1.3545545093390474,
+      "learning_rate": 4.414893617021277e-06,
+      "loss": 1.262,
+      "step": 83
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 1.3348494465098057,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 1.231,
+      "step": 84
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 1.3370363861769372,
+      "learning_rate": 4.521276595744681e-06,
+      "loss": 1.2136,
+      "step": 85
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 1.2502879949642949,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 1.2387,
+      "step": 86
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 1.3634840542204891,
+      "learning_rate": 4.6276595744680855e-06,
+      "loss": 1.2795,
+      "step": 87
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 1.4081794721937446,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.2826,
+      "step": 88
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 1.2169282110477329,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 1.2286,
+      "step": 89
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 1.2421902440104342,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 1.2091,
+      "step": 90
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 1.2215257582048158,
+      "learning_rate": 4.840425531914894e-06,
+      "loss": 1.1016,
+      "step": 91
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 1.2387285211853014,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 1.1803,
+      "step": 92
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 1.263019124972462,
+      "learning_rate": 4.946808510638298e-06,
+      "loss": 1.2403,
+      "step": 93
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 1.2194501954476942,
+      "learning_rate": 5e-06,
+      "loss": 1.2087,
+      "step": 94
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 1.2800554218702347,
+      "learning_rate": 5.053191489361703e-06,
+      "loss": 1.1979,
+      "step": 95
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 1.2503994395846336,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 1.1459,
+      "step": 96
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 1.2648019806917759,
+      "learning_rate": 5.159574468085107e-06,
+      "loss": 1.1908,
+      "step": 97
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 1.2361469771451743,
+      "learning_rate": 5.212765957446809e-06,
+      "loss": 1.175,
+      "step": 98
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 1.2027508478924163,
+      "learning_rate": 5.265957446808511e-06,
+      "loss": 1.1522,
+      "step": 99
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3458900442550417,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 1.2066,
+      "step": 100
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 1.1947554702060192,
+      "learning_rate": 5.372340425531915e-06,
+      "loss": 1.1649,
+      "step": 101
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 1.2814903215405857,
+      "learning_rate": 5.425531914893617e-06,
+      "loss": 1.221,
+      "step": 102
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 1.3228525735133247,
+      "learning_rate": 5.47872340425532e-06,
+      "loss": 1.2608,
+      "step": 103
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.3695791790091256,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 1.3016,
+      "step": 104
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 1.4482515085810126,
+      "learning_rate": 5.5851063829787235e-06,
+      "loss": 1.206,
+      "step": 105
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 1.3572804914450192,
+      "learning_rate": 5.638297872340426e-06,
+      "loss": 1.2277,
+      "step": 106
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 1.345771109352615,
+      "learning_rate": 5.691489361702128e-06,
+      "loss": 1.2686,
+      "step": 107
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 1.353214129611976,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 1.2911,
+      "step": 108
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 1.3544364231107242,
+      "learning_rate": 5.7978723404255325e-06,
+      "loss": 1.2236,
+      "step": 109
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 1.3369920111726017,
+      "learning_rate": 5.851063829787235e-06,
+      "loss": 1.2427,
+      "step": 110
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 1.361279947484296,
+      "learning_rate": 5.904255319148937e-06,
+      "loss": 1.2331,
+      "step": 111
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 1.3405858177565884,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.2419,
+      "step": 112
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 1.3556521536258566,
+      "learning_rate": 6.010638297872341e-06,
+      "loss": 1.1469,
+      "step": 113
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 1.27647979197382,
+      "learning_rate": 6.063829787234044e-06,
+      "loss": 1.2178,
+      "step": 114
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 1.3280288219599166,
+      "learning_rate": 6.117021276595745e-06,
+      "loss": 1.2371,
+      "step": 115
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 1.4032047715117417,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 1.2646,
+      "step": 116
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 1.3875283248481962,
+      "learning_rate": 6.22340425531915e-06,
+      "loss": 1.2687,
+      "step": 117
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 1.3339642396094988,
+      "learning_rate": 6.276595744680851e-06,
+      "loss": 1.2241,
+      "step": 118
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 1.2965813728237912,
+      "learning_rate": 6.329787234042554e-06,
+      "loss": 1.2422,
+      "step": 119
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.232840996946032,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 1.177,
+      "step": 120
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 1.2002854711824298,
+      "learning_rate": 6.436170212765958e-06,
+      "loss": 1.1651,
+      "step": 121
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 1.4467750151272263,
+      "learning_rate": 6.48936170212766e-06,
+      "loss": 1.226,
+      "step": 122
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 1.483516839204314,
+      "learning_rate": 6.542553191489362e-06,
+      "loss": 1.1926,
+      "step": 123
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 1.3636980602642683,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 1.2363,
+      "step": 124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3360595736449652,
+      "learning_rate": 6.648936170212767e-06,
+      "loss": 1.2099,
+      "step": 125
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 1.3202998883065726,
+      "learning_rate": 6.702127659574469e-06,
+      "loss": 1.2568,
+      "step": 126
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 1.4733714471362398,
+      "learning_rate": 6.75531914893617e-06,
+      "loss": 1.2154,
+      "step": 127
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 1.3183132745101613,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.1829,
+      "step": 128
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 1.3396570123523743,
+      "learning_rate": 6.861702127659575e-06,
+      "loss": 1.2214,
+      "step": 129
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.2432539218587322,
+      "learning_rate": 6.914893617021278e-06,
+      "loss": 1.128,
+      "step": 130
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 1.335407974223568,
+      "learning_rate": 6.968085106382979e-06,
+      "loss": 1.2019,
+      "step": 131
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 1.468689388744961,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 1.2138,
+      "step": 132
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 1.4180366341573534,
+      "learning_rate": 7.074468085106384e-06,
+      "loss": 1.1725,
+      "step": 133
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 1.4536203605806246,
+      "learning_rate": 7.127659574468085e-06,
+      "loss": 1.2175,
+      "step": 134
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 1.2644993814847247,
+      "learning_rate": 7.1808510638297875e-06,
+      "loss": 1.206,
+      "step": 135
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.4912773773358907,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 1.2077,
+      "step": 136
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 1.4487467583672629,
+      "learning_rate": 7.287234042553192e-06,
+      "loss": 1.2461,
+      "step": 137
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 1.3514400181572792,
+      "learning_rate": 7.340425531914894e-06,
+      "loss": 1.1677,
+      "step": 138
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 1.4160768735860554,
+      "learning_rate": 7.3936170212765965e-06,
+      "loss": 1.2752,
+      "step": 139
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 1.6465571454842867,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 1.2588,
+      "step": 140
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 1.4947442704576295,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.1323,
+      "step": 141
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 1.4470318856040942,
+      "learning_rate": 7.553191489361703e-06,
+      "loss": 1.2423,
+      "step": 142
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 1.3764661553387543,
+      "learning_rate": 7.606382978723405e-06,
+      "loss": 1.191,
+      "step": 143
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.2865130632592365,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.1355,
+      "step": 144
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.3684410438454009,
+      "learning_rate": 7.71276595744681e-06,
+      "loss": 1.1508,
+      "step": 145
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 1.4825365407894726,
+      "learning_rate": 7.765957446808511e-06,
+      "loss": 1.1853,
+      "step": 146
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 1.427522472786184,
+      "learning_rate": 7.819148936170213e-06,
+      "loss": 1.1631,
+      "step": 147
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 1.3453982047037698,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 1.1864,
+      "step": 148
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 1.3383201271448923,
+      "learning_rate": 7.925531914893617e-06,
+      "loss": 1.2128,
+      "step": 149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2825553959113618,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 1.1445,
+      "step": 150
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 1.4344558305478945,
+      "learning_rate": 8.031914893617022e-06,
+      "loss": 1.2357,
+      "step": 151
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.4613640887135835,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 1.2638,
+      "step": 152
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 1.296519283988,
+      "learning_rate": 8.138297872340426e-06,
+      "loss": 1.1012,
+      "step": 153
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 1.4288650891730967,
+      "learning_rate": 8.191489361702128e-06,
+      "loss": 1.2483,
+      "step": 154
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.4505510144448936,
+      "learning_rate": 8.24468085106383e-06,
+      "loss": 1.2279,
+      "step": 155
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 1.2582211464941604,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 1.1631,
+      "step": 156
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 1.343219139004632,
+      "learning_rate": 8.351063829787235e-06,
+      "loss": 1.1469,
+      "step": 157
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 1.2149488581020176,
+      "learning_rate": 8.404255319148937e-06,
+      "loss": 1.0701,
+      "step": 158
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 1.4844514577515964,
+      "learning_rate": 8.457446808510638e-06,
+      "loss": 1.275,
+      "step": 159
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.3963129808729753,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 1.1679,
+      "step": 160
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 1.4011070218880282,
+      "learning_rate": 8.563829787234044e-06,
+      "loss": 1.187,
+      "step": 161
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.4460463927692926,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 1.188,
+      "step": 162
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 1.2744963443999946,
+      "learning_rate": 8.670212765957447e-06,
+      "loss": 1.1603,
+      "step": 163
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 1.3360709566920668,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 1.1424,
+      "step": 164
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.3581704195355544,
+      "learning_rate": 8.776595744680852e-06,
+      "loss": 1.2209,
+      "step": 165
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 1.2772659444525245,
+      "learning_rate": 8.829787234042555e-06,
+      "loss": 1.1161,
+      "step": 166
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 1.3146686211403726,
+      "learning_rate": 8.882978723404256e-06,
+      "loss": 1.1196,
+      "step": 167
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.3394702888408314,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.218,
+      "step": 168
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 1.497676994277907,
+      "learning_rate": 8.98936170212766e-06,
+      "loss": 1.1524,
+      "step": 169
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.571636390734505,
+      "learning_rate": 9.042553191489362e-06,
+      "loss": 1.261,
+      "step": 170
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 1.3814744647519746,
+      "learning_rate": 9.095744680851063e-06,
+      "loss": 1.1747,
+      "step": 171
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 1.37492407457705,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 1.1898,
+      "step": 172
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 1.4416592772050616,
+      "learning_rate": 9.20212765957447e-06,
+      "loss": 1.1893,
+      "step": 173
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 1.4801710265103374,
+      "learning_rate": 9.255319148936171e-06,
+      "loss": 1.1473,
+      "step": 174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4337770310627218,
+      "learning_rate": 9.308510638297872e-06,
+      "loss": 1.1725,
+      "step": 175
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.2891377685256382,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 1.182,
+      "step": 176
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 1.335645497990819,
+      "learning_rate": 9.414893617021279e-06,
+      "loss": 1.2227,
+      "step": 177
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 1.3081578576211257,
+      "learning_rate": 9.46808510638298e-06,
+      "loss": 1.1819,
+      "step": 178
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 1.301421253758618,
+      "learning_rate": 9.521276595744681e-06,
+      "loss": 1.1593,
+      "step": 179
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.3337804593717932,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 1.0941,
+      "step": 180
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 1.3772510705384873,
+      "learning_rate": 9.627659574468086e-06,
+      "loss": 1.1978,
+      "step": 181
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 1.2867843810362423,
+      "learning_rate": 9.680851063829787e-06,
+      "loss": 1.074,
+      "step": 182
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 1.3566501945984397,
+      "learning_rate": 9.73404255319149e-06,
+      "loss": 1.1557,
+      "step": 183
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 1.3032741383015087,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 1.1435,
+      "step": 184
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.521106793220149,
+      "learning_rate": 9.840425531914895e-06,
+      "loss": 1.1413,
+      "step": 185
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 1.3989911756975302,
+      "learning_rate": 9.893617021276596e-06,
+      "loss": 1.1117,
+      "step": 186
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 1.2035415678694272,
+      "learning_rate": 9.946808510638298e-06,
+      "loss": 1.1227,
+      "step": 187
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 1.5488537450625632,
+      "learning_rate": 1e-05,
+      "loss": 1.1624,
+      "step": 188
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 1.3700674838059985,
+      "learning_rate": 9.99999133019233e-06,
+      "loss": 1.1637,
+      "step": 189
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.5349319366797567,
+      "learning_rate": 9.999965320799377e-06,
+      "loss": 1.1556,
+      "step": 190
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 1.6125124209777197,
+      "learning_rate": 9.999921971911345e-06,
+      "loss": 1.2168,
+      "step": 191
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 1.6566049406398442,
+      "learning_rate": 9.999861283678563e-06,
+      "loss": 1.2343,
+      "step": 192
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 1.6408125099408117,
+      "learning_rate": 9.999783256311494e-06,
+      "loss": 1.1576,
+      "step": 193
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 1.740155138832122,
+      "learning_rate": 9.99968789008073e-06,
+      "loss": 1.1909,
+      "step": 194
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.3601981225220867,
+      "learning_rate": 9.999575185316994e-06,
+      "loss": 1.1268,
+      "step": 195
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 1.431564889308234,
+      "learning_rate": 9.999445142411139e-06,
+      "loss": 1.1545,
+      "step": 196
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 1.5350959943206477,
+      "learning_rate": 9.99929776181414e-06,
+      "loss": 1.1764,
+      "step": 197
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 1.4432635428618188,
+      "learning_rate": 9.999133044037107e-06,
+      "loss": 1.1307,
+      "step": 198
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 1.474178678764435,
+      "learning_rate": 9.998950989651261e-06,
+      "loss": 1.2215,
+      "step": 199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5422938099960322,
+      "learning_rate": 9.99875159928796e-06,
+      "loss": 1.2116,
+      "step": 200
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 1.6010305201244435,
+      "learning_rate": 9.99853487363867e-06,
+      "loss": 1.1784,
+      "step": 201
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 1.4799525806803449,
+      "learning_rate": 9.998300813454981e-06,
+      "loss": 1.176,
+      "step": 202
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 1.24248095082439,
+      "learning_rate": 9.998049419548597e-06,
+      "loss": 1.0655,
+      "step": 203
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 1.3237786767118838,
+      "learning_rate": 9.997780692791329e-06,
+      "loss": 1.1385,
+      "step": 204
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 1.376207748009853,
+      "learning_rate": 9.997494634115101e-06,
+      "loss": 1.1771,
+      "step": 205
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 1.51037915203655,
+      "learning_rate": 9.997191244511947e-06,
+      "loss": 1.1682,
+      "step": 206
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 1.466680370898261,
+      "learning_rate": 9.996870525033994e-06,
+      "loss": 1.2051,
+      "step": 207
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 1.3634930294008605,
+      "learning_rate": 9.996532476793475e-06,
+      "loss": 1.1202,
+      "step": 208
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 1.2465835964756378,
+      "learning_rate": 9.996177100962714e-06,
+      "loss": 1.1694,
+      "step": 209
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.3874386369129013,
+      "learning_rate": 9.995804398774129e-06,
+      "loss": 1.0901,
+      "step": 210
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 1.345360982064232,
+      "learning_rate": 9.99541437152022e-06,
+      "loss": 1.1712,
+      "step": 211
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 1.419035335386871,
+      "learning_rate": 9.995007020553572e-06,
+      "loss": 1.1157,
+      "step": 212
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 1.3209322815554059,
+      "learning_rate": 9.994582347286849e-06,
+      "loss": 1.0652,
+      "step": 213
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 1.3236849677598421,
+      "learning_rate": 9.994140353192782e-06,
+      "loss": 1.1161,
+      "step": 214
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.5461439893998776,
+      "learning_rate": 9.993681039804176e-06,
+      "loss": 1.2206,
+      "step": 215
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 1.2659730424463869,
+      "learning_rate": 9.99320440871389e-06,
+      "loss": 1.124,
+      "step": 216
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 1.2952391235729264,
+      "learning_rate": 9.99271046157485e-06,
+      "loss": 1.1312,
+      "step": 217
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 1.3921954065681275,
+      "learning_rate": 9.99219920010002e-06,
+      "loss": 1.1248,
+      "step": 218
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 1.4079765768683832,
+      "learning_rate": 9.991670626062422e-06,
+      "loss": 1.0761,
+      "step": 219
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.4672083065816608,
+      "learning_rate": 9.991124741295106e-06,
+      "loss": 1.1854,
+      "step": 220
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 1.5647122615781168,
+      "learning_rate": 9.990561547691159e-06,
+      "loss": 1.195,
+      "step": 221
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 1.4841111563489182,
+      "learning_rate": 9.989981047203693e-06,
+      "loss": 1.0481,
+      "step": 222
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 1.3561828466541594,
+      "learning_rate": 9.98938324184584e-06,
+      "loss": 1.0719,
+      "step": 223
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 1.4316308480057511,
+      "learning_rate": 9.988768133690741e-06,
+      "loss": 1.2046,
+      "step": 224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.229225422203295,
+      "learning_rate": 9.988135724871546e-06,
+      "loss": 1.0545,
+      "step": 225
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 1.5102880958499596,
+      "learning_rate": 9.987486017581401e-06,
+      "loss": 1.1112,
+      "step": 226
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 1.3961232122907166,
+      "learning_rate": 9.986819014073436e-06,
+      "loss": 1.1292,
+      "step": 227
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 1.3891393586296323,
+      "learning_rate": 9.986134716660774e-06,
+      "loss": 1.0928,
+      "step": 228
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 1.4710516139625378,
+      "learning_rate": 9.9854331277165e-06,
+      "loss": 1.1266,
+      "step": 229
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 1.3350749945807487,
+      "learning_rate": 9.984714249673676e-06,
+      "loss": 1.0565,
+      "step": 230
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 1.401738250670105,
+      "learning_rate": 9.98397808502531e-06,
+      "loss": 1.1217,
+      "step": 231
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 1.2886111933313193,
+      "learning_rate": 9.983224636324369e-06,
+      "loss": 1.0816,
+      "step": 232
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 1.436954677162413,
+      "learning_rate": 9.982453906183754e-06,
+      "loss": 1.2147,
+      "step": 233
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 1.5126390843338893,
+      "learning_rate": 9.981665897276298e-06,
+      "loss": 1.0673,
+      "step": 234
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 1.5316367681014587,
+      "learning_rate": 9.980860612334753e-06,
+      "loss": 1.1157,
+      "step": 235
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 1.4224539271856318,
+      "learning_rate": 9.980038054151789e-06,
+      "loss": 1.1385,
+      "step": 236
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 1.530543816532965,
+      "learning_rate": 9.979198225579968e-06,
+      "loss": 1.1508,
+      "step": 237
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 1.5108829686404601,
+      "learning_rate": 9.97834112953176e-06,
+      "loss": 1.1154,
+      "step": 238
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 1.4564345557774694,
+      "learning_rate": 9.9774667689795e-06,
+      "loss": 1.1161,
+      "step": 239
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 1.3290973902900243,
+      "learning_rate": 9.976575146955409e-06,
+      "loss": 1.0322,
+      "step": 240
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 1.2659088125835956,
+      "learning_rate": 9.97566626655156e-06,
+      "loss": 1.0965,
+      "step": 241
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 1.4192231711419405,
+      "learning_rate": 9.974740130919883e-06,
+      "loss": 1.0634,
+      "step": 242
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 1.508350949154953,
+      "learning_rate": 9.973796743272141e-06,
+      "loss": 1.0853,
+      "step": 243
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 1.4712841495273241,
+      "learning_rate": 9.972836106879936e-06,
+      "loss": 1.0686,
+      "step": 244
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 1.328872572714548,
+      "learning_rate": 9.971858225074672e-06,
+      "loss": 1.0567,
+      "step": 245
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 1.397809579235734,
+      "learning_rate": 9.970863101247578e-06,
+      "loss": 1.1113,
+      "step": 246
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 1.490591810704279,
+      "learning_rate": 9.96985073884966e-06,
+      "loss": 1.1524,
+      "step": 247
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 1.624262558518088,
+      "learning_rate": 9.968821141391716e-06,
+      "loss": 1.1729,
+      "step": 248
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 1.427929092457274,
+      "learning_rate": 9.96777431244431e-06,
+      "loss": 1.1092,
+      "step": 249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4009366097525648,
+      "learning_rate": 9.966710255637764e-06,
+      "loss": 1.1142,
+      "step": 250
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 1.3619657069373443,
+      "learning_rate": 9.965628974662145e-06,
+      "loss": 1.1196,
+      "step": 251
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 1.286213190237012,
+      "learning_rate": 9.964530473267253e-06,
+      "loss": 1.1211,
+      "step": 252
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 1.348982327724463,
+      "learning_rate": 9.963414755262606e-06,
+      "loss": 1.0638,
+      "step": 253
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 1.3198502705604718,
+      "learning_rate": 9.962281824517427e-06,
+      "loss": 1.1044,
+      "step": 254
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 1.295208603915589,
+      "learning_rate": 9.961131684960635e-06,
+      "loss": 1.0912,
+      "step": 255
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 1.316775511589202,
+      "learning_rate": 9.959964340580823e-06,
+      "loss": 1.1581,
+      "step": 256
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 1.459171866725438,
+      "learning_rate": 9.958779795426253e-06,
+      "loss": 1.2307,
+      "step": 257
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 2.071160152062741,
+      "learning_rate": 9.957578053604837e-06,
+      "loss": 1.1221,
+      "step": 258
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 1.3753345270435529,
+      "learning_rate": 9.956359119284123e-06,
+      "loss": 1.0596,
+      "step": 259
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 1.4079035619067826,
+      "learning_rate": 9.955122996691278e-06,
+      "loss": 1.1707,
+      "step": 260
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 1.2988760331074598,
+      "learning_rate": 9.953869690113085e-06,
+      "loss": 1.054,
+      "step": 261
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 1.2299978341794726,
+      "learning_rate": 9.952599203895912e-06,
+      "loss": 0.964,
+      "step": 262
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 1.2675070599935245,
+      "learning_rate": 9.95131154244571e-06,
+      "loss": 1.1555,
+      "step": 263
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 1.3899921710506915,
+      "learning_rate": 9.950006710227986e-06,
+      "loss": 1.1785,
+      "step": 264
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 1.6143530294905075,
+      "learning_rate": 9.9486847117678e-06,
+      "loss": 1.1335,
+      "step": 265
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 1.407531402482802,
+      "learning_rate": 9.947345551649741e-06,
+      "loss": 1.1162,
+      "step": 266
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 1.3379337446628576,
+      "learning_rate": 9.945989234517913e-06,
+      "loss": 1.1108,
+      "step": 267
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 1.4199178880182914,
+      "learning_rate": 9.94461576507592e-06,
+      "loss": 1.0977,
+      "step": 268
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 1.2703762317921958,
+      "learning_rate": 9.943225148086846e-06,
+      "loss": 1.0604,
+      "step": 269
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.4444914487644183,
+      "learning_rate": 9.941817388373248e-06,
+      "loss": 1.0586,
+      "step": 270
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 1.423934584606783,
+      "learning_rate": 9.940392490817124e-06,
+      "loss": 1.1046,
+      "step": 271
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 1.371960653254297,
+      "learning_rate": 9.938950460359912e-06,
+      "loss": 1.1095,
+      "step": 272
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 1.3457706855286338,
+      "learning_rate": 9.937491302002462e-06,
+      "loss": 1.092,
+      "step": 273
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 1.2968675257429614,
+      "learning_rate": 9.936015020805022e-06,
+      "loss": 1.058,
+      "step": 274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.356124443843827,
+      "learning_rate": 9.934521621887223e-06,
+      "loss": 1.1312,
+      "step": 275
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 1.3896576115432853,
+      "learning_rate": 9.933011110428058e-06,
+      "loss": 1.0579,
+      "step": 276
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 1.298966949578673,
+      "learning_rate": 9.93148349166586e-06,
+      "loss": 1.0388,
+      "step": 277
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 1.3704742845904228,
+      "learning_rate": 9.929938770898299e-06,
+      "loss": 1.093,
+      "step": 278
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 1.5321224395563418,
+      "learning_rate": 9.928376953482343e-06,
+      "loss": 1.0876,
+      "step": 279
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 1.4173356406870716,
+      "learning_rate": 9.926798044834261e-06,
+      "loss": 1.0978,
+      "step": 280
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 1.5355690456515665,
+      "learning_rate": 9.92520205042958e-06,
+      "loss": 1.2086,
+      "step": 281
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 1.5686218515796675,
+      "learning_rate": 9.92358897580309e-06,
+      "loss": 1.1348,
+      "step": 282
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 1.3614651917760048,
+      "learning_rate": 9.921958826548808e-06,
+      "loss": 1.0557,
+      "step": 283
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 1.4808899417219499,
+      "learning_rate": 9.920311608319968e-06,
+      "loss": 1.1303,
+      "step": 284
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 1.4088018049846256,
+      "learning_rate": 9.918647326828993e-06,
+      "loss": 1.1438,
+      "step": 285
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 1.4234523935973298,
+      "learning_rate": 9.916965987847485e-06,
+      "loss": 1.108,
+      "step": 286
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 1.3083195776197443,
+      "learning_rate": 9.915267597206198e-06,
+      "loss": 0.9485,
+      "step": 287
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 1.419564765988596,
+      "learning_rate": 9.913552160795022e-06,
+      "loss": 1.0833,
+      "step": 288
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 1.4349711103964689,
+      "learning_rate": 9.911819684562954e-06,
+      "loss": 1.1133,
+      "step": 289
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 1.41655866439432,
+      "learning_rate": 9.910070174518093e-06,
+      "loss": 1.0961,
+      "step": 290
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 1.2677493706871048,
+      "learning_rate": 9.908303636727604e-06,
+      "loss": 1.0251,
+      "step": 291
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 1.3501832149488533,
+      "learning_rate": 9.9065200773177e-06,
+      "loss": 1.083,
+      "step": 292
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 1.460231278456003,
+      "learning_rate": 9.904719502473635e-06,
+      "loss": 1.1137,
+      "step": 293
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 1.3036411536339068,
+      "learning_rate": 9.902901918439658e-06,
+      "loss": 0.9734,
+      "step": 294
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 1.5395605105559098,
+      "learning_rate": 9.901067331519013e-06,
+      "loss": 1.0424,
+      "step": 295
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 1.4085757683413787,
+      "learning_rate": 9.899215748073906e-06,
+      "loss": 1.0396,
+      "step": 296
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 1.4537306617533265,
+      "learning_rate": 9.897347174525487e-06,
+      "loss": 1.1607,
+      "step": 297
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 1.3841071230965598,
+      "learning_rate": 9.895461617353823e-06,
+      "loss": 1.0525,
+      "step": 298
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 1.3868983041275091,
+      "learning_rate": 9.893559083097885e-06,
+      "loss": 1.0444,
+      "step": 299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4568066176781722,
+      "learning_rate": 9.891639578355511e-06,
+      "loss": 1.0736,
+      "step": 300
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 1.3067515068359825,
+      "learning_rate": 9.8897031097834e-06,
+      "loss": 1.1044,
+      "step": 301
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 1.4260807731079035,
+      "learning_rate": 9.887749684097072e-06,
+      "loss": 1.0953,
+      "step": 302
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 1.80592454903872,
+      "learning_rate": 9.88577930807086e-06,
+      "loss": 1.1976,
+      "step": 303
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 1.518550004472384,
+      "learning_rate": 9.883791988537874e-06,
+      "loss": 1.1057,
+      "step": 304
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 1.4266574785527073,
+      "learning_rate": 9.881787732389987e-06,
+      "loss": 1.0799,
+      "step": 305
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 1.356443023598987,
+      "learning_rate": 9.879766546577805e-06,
+      "loss": 1.1128,
+      "step": 306
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 1.4720434171341803,
+      "learning_rate": 9.877728438110645e-06,
+      "loss": 1.1867,
+      "step": 307
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 1.4448616854522536,
+      "learning_rate": 9.87567341405651e-06,
+      "loss": 1.0513,
+      "step": 308
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 1.5292387427331948,
+      "learning_rate": 9.873601481542065e-06,
+      "loss": 1.1308,
+      "step": 309
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 1.5960318173461212,
+      "learning_rate": 9.871512647752612e-06,
+      "loss": 1.0847,
+      "step": 310
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 1.4239040091073365,
+      "learning_rate": 9.86940691993207e-06,
+      "loss": 1.1631,
+      "step": 311
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 1.3617595889098688,
+      "learning_rate": 9.867284305382936e-06,
+      "loss": 1.1187,
+      "step": 312
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 1.5144990765622857,
+      "learning_rate": 9.865144811466275e-06,
+      "loss": 1.2349,
+      "step": 313
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 1.4413232026956877,
+      "learning_rate": 9.86298844560169e-06,
+      "loss": 1.2465,
+      "step": 314
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 1.376768835793776,
+      "learning_rate": 9.860815215267288e-06,
+      "loss": 1.0037,
+      "step": 315
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 1.4503533577704,
+      "learning_rate": 9.858625127999668e-06,
+      "loss": 1.11,
+      "step": 316
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 1.2915101862696745,
+      "learning_rate": 9.856418191393881e-06,
+      "loss": 1.1137,
+      "step": 317
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 1.4236402213132213,
+      "learning_rate": 9.854194413103418e-06,
+      "loss": 1.165,
+      "step": 318
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 1.3333599162084,
+      "learning_rate": 9.851953800840166e-06,
+      "loss": 1.0299,
+      "step": 319
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 1.5293465874553285,
+      "learning_rate": 9.849696362374399e-06,
+      "loss": 1.1515,
+      "step": 320
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 1.2696937195306555,
+      "learning_rate": 9.847422105534739e-06,
+      "loss": 0.996,
+      "step": 321
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 1.4914577758056384,
+      "learning_rate": 9.845131038208135e-06,
+      "loss": 1.0589,
+      "step": 322
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 1.3309646892339204,
+      "learning_rate": 9.84282316833983e-06,
+      "loss": 1.1223,
+      "step": 323
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 1.6547862080564029,
+      "learning_rate": 9.84049850393334e-06,
+      "loss": 1.1241,
+      "step": 324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4446068581433718,
+      "learning_rate": 9.838157053050423e-06,
+      "loss": 1.1515,
+      "step": 325
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 1.3572902614510736,
+      "learning_rate": 9.83579882381105e-06,
+      "loss": 0.9326,
+      "step": 326
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 1.3032141638666164,
+      "learning_rate": 9.83342382439338e-06,
+      "loss": 0.9887,
+      "step": 327
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 1.476111488230785,
+      "learning_rate": 9.831032063033726e-06,
+      "loss": 1.0604,
+      "step": 328
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 1.293357383036506,
+      "learning_rate": 9.828623548026533e-06,
+      "loss": 1.0694,
+      "step": 329
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 1.3393010031541832,
+      "learning_rate": 9.826198287724346e-06,
+      "loss": 1.0594,
+      "step": 330
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 1.5578323587269973,
+      "learning_rate": 9.823756290537783e-06,
+      "loss": 1.0911,
+      "step": 331
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 1.4727909180652679,
+      "learning_rate": 9.821297564935499e-06,
+      "loss": 1.0078,
+      "step": 332
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 1.4156377789684005,
+      "learning_rate": 9.81882211944417e-06,
+      "loss": 0.9984,
+      "step": 333
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 1.425316618473854,
+      "learning_rate": 9.816329962648444e-06,
+      "loss": 1.0666,
+      "step": 334
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 1.4749573178089799,
+      "learning_rate": 9.813821103190932e-06,
+      "loss": 1.042,
+      "step": 335
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 1.466236786142977,
+      "learning_rate": 9.811295549772169e-06,
+      "loss": 1.0649,
+      "step": 336
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 1.4394346664234121,
+      "learning_rate": 9.808753311150575e-06,
+      "loss": 1.0583,
+      "step": 337
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 1.4266866659060504,
+      "learning_rate": 9.80619439614244e-06,
+      "loss": 0.927,
+      "step": 338
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 1.4613623808585878,
+      "learning_rate": 9.803618813621885e-06,
+      "loss": 1.0895,
+      "step": 339
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.2752317854741668,
+      "learning_rate": 9.801026572520832e-06,
+      "loss": 0.9663,
+      "step": 340
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 1.2578240474978186,
+      "learning_rate": 9.798417681828972e-06,
+      "loss": 1.0611,
+      "step": 341
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 1.3796245139401346,
+      "learning_rate": 9.795792150593739e-06,
+      "loss": 1.0326,
+      "step": 342
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 1.4335318454737809,
+      "learning_rate": 9.793149987920273e-06,
+      "loss": 1.1053,
+      "step": 343
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 1.3198025876924644,
+      "learning_rate": 9.79049120297139e-06,
+      "loss": 1.0434,
+      "step": 344
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 1.4272450602656777,
+      "learning_rate": 9.787815804967554e-06,
+      "loss": 1.0491,
+      "step": 345
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 1.3981670027118367,
+      "learning_rate": 9.785123803186834e-06,
+      "loss": 0.9857,
+      "step": 346
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 1.3368617371666986,
+      "learning_rate": 9.782415206964892e-06,
+      "loss": 1.1536,
+      "step": 347
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 1.4323567295548643,
+      "learning_rate": 9.779690025694926e-06,
+      "loss": 1.0893,
+      "step": 348
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 1.399485624494657,
+      "learning_rate": 9.776948268827658e-06,
+      "loss": 1.0643,
+      "step": 349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3484818742842373,
+      "learning_rate": 9.77418994587129e-06,
+      "loss": 0.9572,
+      "step": 350
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 1.4289747261006025,
+      "learning_rate": 9.771415066391473e-06,
+      "loss": 1.1299,
+      "step": 351
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 1.5667382297161747,
+      "learning_rate": 9.768623640011272e-06,
+      "loss": 1.0719,
+      "step": 352
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 1.5540796356790698,
+      "learning_rate": 9.765815676411145e-06,
+      "loss": 1.0784,
+      "step": 353
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 1.3587673433045933,
+      "learning_rate": 9.762991185328891e-06,
+      "loss": 1.0499,
+      "step": 354
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 1.4033939019167627,
+      "learning_rate": 9.760150176559627e-06,
+      "loss": 1.0151,
+      "step": 355
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 1.4049000175769197,
+      "learning_rate": 9.757292659955755e-06,
+      "loss": 1.0542,
+      "step": 356
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 1.3434493531416045,
+      "learning_rate": 9.754418645426919e-06,
+      "loss": 0.9762,
+      "step": 357
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 1.3616889165101278,
+      "learning_rate": 9.751528142939986e-06,
+      "loss": 0.9538,
+      "step": 358
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 1.4020696983713603,
+      "learning_rate": 9.74862116251899e-06,
+      "loss": 1.1129,
+      "step": 359
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 1.3827614900673553,
+      "learning_rate": 9.74569771424512e-06,
+      "loss": 0.9426,
+      "step": 360
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 1.4792586595847994,
+      "learning_rate": 9.742757808256667e-06,
+      "loss": 1.1004,
+      "step": 361
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 1.4224067433300234,
+      "learning_rate": 9.739801454749e-06,
+      "loss": 1.0168,
+      "step": 362
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 1.566855307680239,
+      "learning_rate": 9.736828663974527e-06,
+      "loss": 1.0908,
+      "step": 363
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 1.363750654977135,
+      "learning_rate": 9.733839446242655e-06,
+      "loss": 1.0617,
+      "step": 364
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 1.2351780465049123,
+      "learning_rate": 9.730833811919763e-06,
+      "loss": 0.9059,
+      "step": 365
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 1.6293002889532675,
+      "learning_rate": 9.727811771429158e-06,
+      "loss": 0.9951,
+      "step": 366
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 1.5305070627940711,
+      "learning_rate": 9.724773335251046e-06,
+      "loss": 1.176,
+      "step": 367
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 1.3277686876255392,
+      "learning_rate": 9.721718513922488e-06,
+      "loss": 1.0177,
+      "step": 368
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 1.356984668070833,
+      "learning_rate": 9.71864731803737e-06,
+      "loss": 1.0318,
+      "step": 369
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 1.2683522694587366,
+      "learning_rate": 9.715559758246363e-06,
+      "loss": 1.0251,
+      "step": 370
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 1.424444426436754,
+      "learning_rate": 9.712455845256888e-06,
+      "loss": 1.098,
+      "step": 371
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 1.3341358167076778,
+      "learning_rate": 9.709335589833076e-06,
+      "loss": 1.1373,
+      "step": 372
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 1.2870327681580784,
+      "learning_rate": 9.70619900279573e-06,
+      "loss": 0.8975,
+      "step": 373
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 1.3348420386376083,
+      "learning_rate": 9.703046095022297e-06,
+      "loss": 1.0064,
+      "step": 374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3709173376168697,
+      "learning_rate": 9.699876877446815e-06,
+      "loss": 1.1438,
+      "step": 375
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 1.3760720157867286,
+      "learning_rate": 9.696691361059886e-06,
+      "loss": 1.0253,
+      "step": 376
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 1.5382638334918615,
+      "learning_rate": 9.693489556908641e-06,
+      "loss": 1.037,
+      "step": 377
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 1.351490372894956,
+      "learning_rate": 9.690271476096686e-06,
+      "loss": 1.16,
+      "step": 378
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 1.365510502558049,
+      "learning_rate": 9.68703712978408e-06,
+      "loss": 1.0615,
+      "step": 379
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 1.3099829805135041,
+      "learning_rate": 9.683786529187287e-06,
+      "loss": 1.0203,
+      "step": 380
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 1.3157866637545363,
+      "learning_rate": 9.680519685579137e-06,
+      "loss": 1.056,
+      "step": 381
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 1.3650624681981065,
+      "learning_rate": 9.677236610288797e-06,
+      "loss": 0.9898,
+      "step": 382
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 1.3848893593322975,
+      "learning_rate": 9.673937314701714e-06,
+      "loss": 0.948,
+      "step": 383
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.3497879323322155,
+      "learning_rate": 9.670621810259596e-06,
+      "loss": 1.0288,
+      "step": 384
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 1.565925882114841,
+      "learning_rate": 9.667290108460354e-06,
+      "loss": 0.972,
+      "step": 385
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 1.3291071954446392,
+      "learning_rate": 9.663942220858075e-06,
+      "loss": 0.9313,
+      "step": 386
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 1.4372820942877105,
+      "learning_rate": 9.660578159062977e-06,
+      "loss": 0.9868,
+      "step": 387
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 1.5008546489944374,
+      "learning_rate": 9.657197934741366e-06,
+      "loss": 1.1115,
+      "step": 388
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 1.2861900800211155,
+      "learning_rate": 9.6538015596156e-06,
+      "loss": 0.9564,
+      "step": 389
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 1.4194164988442,
+      "learning_rate": 9.650389045464046e-06,
+      "loss": 1.0863,
+      "step": 390
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 1.7349864602957699,
+      "learning_rate": 9.646960404121042e-06,
+      "loss": 1.0481,
+      "step": 391
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 1.61642313812224,
+      "learning_rate": 9.643515647476851e-06,
+      "loss": 1.0813,
+      "step": 392
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 1.3380108321642106,
+      "learning_rate": 9.640054787477626e-06,
+      "loss": 1.0212,
+      "step": 393
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 1.2843270563012295,
+      "learning_rate": 9.63657783612536e-06,
+      "loss": 1.0151,
+      "step": 394
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 1.4779743766918632,
+      "learning_rate": 9.633084805477857e-06,
+      "loss": 1.0835,
+      "step": 395
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 1.5792056912591286,
+      "learning_rate": 9.629575707648675e-06,
+      "loss": 1.0966,
+      "step": 396
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 1.4500281821713112,
+      "learning_rate": 9.626050554807096e-06,
+      "loss": 1.0102,
+      "step": 397
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 1.3757711053855997,
+      "learning_rate": 9.62250935917808e-06,
+      "loss": 1.0563,
+      "step": 398
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 1.3243322100582406,
+      "learning_rate": 9.618952133042223e-06,
+      "loss": 1.0321,
+      "step": 399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3851560486813863,
+      "learning_rate": 9.615378888735706e-06,
+      "loss": 0.9338,
+      "step": 400
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 1.3327827962555758,
+      "learning_rate": 9.611789638650269e-06,
+      "loss": 1.0918,
+      "step": 401
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 1.275686910009996,
+      "learning_rate": 9.608184395233156e-06,
+      "loss": 1.0041,
+      "step": 402
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 1.4781053932291743,
+      "learning_rate": 9.604563170987072e-06,
+      "loss": 0.9842,
+      "step": 403
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 1.350356532097954,
+      "learning_rate": 9.600925978470143e-06,
+      "loss": 1.1321,
+      "step": 404
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 1.3183676957465034,
+      "learning_rate": 9.597272830295877e-06,
+      "loss": 0.8684,
+      "step": 405
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 1.4067556519942528,
+      "learning_rate": 9.593603739133105e-06,
+      "loss": 0.9694,
+      "step": 406
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 1.4033440320123447,
+      "learning_rate": 9.589918717705957e-06,
+      "loss": 1.0279,
+      "step": 407
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 1.3560259043300673,
+      "learning_rate": 9.586217778793804e-06,
+      "loss": 0.9628,
+      "step": 408
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 1.4521228304296208,
+      "learning_rate": 9.582500935231215e-06,
+      "loss": 1.0583,
+      "step": 409
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 1.4615181270645066,
+      "learning_rate": 9.57876819990792e-06,
+      "loss": 1.0718,
+      "step": 410
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 1.3445587241062362,
+      "learning_rate": 9.575019585768758e-06,
+      "loss": 1.0094,
+      "step": 411
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 1.3465793558230457,
+      "learning_rate": 9.571255105813632e-06,
+      "loss": 0.9939,
+      "step": 412
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 1.3068339329954002,
+      "learning_rate": 9.567474773097469e-06,
+      "loss": 0.9838,
+      "step": 413
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 1.3348153035667367,
+      "learning_rate": 9.563678600730175e-06,
+      "loss": 0.9961,
+      "step": 414
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 1.3869880627035986,
+      "learning_rate": 9.559866601876581e-06,
+      "loss": 1.0584,
+      "step": 415
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 1.4161734094656369,
+      "learning_rate": 9.556038789756407e-06,
+      "loss": 0.9802,
+      "step": 416
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 1.403330196693304,
+      "learning_rate": 9.55219517764421e-06,
+      "loss": 0.9121,
+      "step": 417
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 1.375198033781532,
+      "learning_rate": 9.548335778869342e-06,
+      "loss": 1.0393,
+      "step": 418
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 1.4535458138738064,
+      "learning_rate": 9.544460606815901e-06,
+      "loss": 1.0355,
+      "step": 419
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 1.4666829475111294,
+      "learning_rate": 9.540569674922685e-06,
+      "loss": 1.0592,
+      "step": 420
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 1.5661419472529932,
+      "learning_rate": 9.536662996683146e-06,
+      "loss": 1.063,
+      "step": 421
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 1.3649442522240554,
+      "learning_rate": 9.532740585645346e-06,
+      "loss": 1.0373,
+      "step": 422
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 1.522077045841566,
+      "learning_rate": 9.528802455411902e-06,
+      "loss": 1.0959,
+      "step": 423
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 1.3376725840655923,
+      "learning_rate": 9.52484861963995e-06,
+      "loss": 0.9812,
+      "step": 424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5073458649838487,
+      "learning_rate": 9.520879092041085e-06,
+      "loss": 1.0158,
+      "step": 425
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 1.4229412458172932,
+      "learning_rate": 9.516893886381324e-06,
+      "loss": 1.0291,
+      "step": 426
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 1.416849856234234,
+      "learning_rate": 9.512893016481053e-06,
+      "loss": 1.0721,
+      "step": 427
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 1.3025128552242944,
+      "learning_rate": 9.508876496214983e-06,
+      "loss": 1.0327,
+      "step": 428
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 1.2684430584973059,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 1.0313,
+      "step": 429
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 1.4524776583764727,
+      "learning_rate": 9.500796560355603e-06,
+      "loss": 0.9196,
+      "step": 430
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 1.3623624958985843,
+      "learning_rate": 9.496733172782889e-06,
+      "loss": 0.9745,
+      "step": 431
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 1.4941557995765784,
+      "learning_rate": 9.492654190885469e-06,
+      "loss": 1.1198,
+      "step": 432
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 1.454473352918338,
+      "learning_rate": 9.488559628808939e-06,
+      "loss": 0.9703,
+      "step": 433
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 1.5181228433567695,
+      "learning_rate": 9.484449500752927e-06,
+      "loss": 1.0861,
+      "step": 434
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 1.440338143669436,
+      "learning_rate": 9.480323820971039e-06,
+      "loss": 1.0455,
+      "step": 435
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 1.3839783304870064,
+      "learning_rate": 9.476182603770814e-06,
+      "loss": 1.0234,
+      "step": 436
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 1.4315106406727882,
+      "learning_rate": 9.472025863513676e-06,
+      "loss": 0.9373,
+      "step": 437
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 1.460124969435515,
+      "learning_rate": 9.467853614614883e-06,
+      "loss": 0.8719,
+      "step": 438
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 1.3605654972387065,
+      "learning_rate": 9.46366587154347e-06,
+      "loss": 0.9736,
+      "step": 439
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 1.4398229341498925,
+      "learning_rate": 9.459462648822209e-06,
+      "loss": 0.9944,
+      "step": 440
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 1.3425147616546218,
+      "learning_rate": 9.45524396102755e-06,
+      "loss": 0.8649,
+      "step": 441
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 1.3814499977284498,
+      "learning_rate": 9.451009822789583e-06,
+      "loss": 1.0668,
+      "step": 442
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 1.515868431752213,
+      "learning_rate": 9.44676024879197e-06,
+      "loss": 0.9735,
+      "step": 443
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 1.4122398644232614,
+      "learning_rate": 9.442495253771909e-06,
+      "loss": 1.0488,
+      "step": 444
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 1.448253262824963,
+      "learning_rate": 9.438214852520073e-06,
+      "loss": 1.0108,
+      "step": 445
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 1.3898433869765308,
+      "learning_rate": 9.433919059880564e-06,
+      "loss": 1.0446,
+      "step": 446
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 1.4189149610925298,
+      "learning_rate": 9.429607890750863e-06,
+      "loss": 1.078,
+      "step": 447
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 1.454146339770495,
+      "learning_rate": 9.425281360081769e-06,
+      "loss": 0.8818,
+      "step": 448
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 1.274658994708157,
+      "learning_rate": 9.420939482877359e-06,
+      "loss": 0.8326,
+      "step": 449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4398693953921002,
+      "learning_rate": 9.416582274194929e-06,
+      "loss": 0.9421,
+      "step": 450
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 1.4731642102431786,
+      "learning_rate": 9.412209749144947e-06,
+      "loss": 0.9234,
+      "step": 451
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 1.5352033907226434,
+      "learning_rate": 9.40782192289099e-06,
+      "loss": 0.8116,
+      "step": 452
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 1.4023895783840634,
+      "learning_rate": 9.4034188106497e-06,
+      "loss": 1.0467,
+      "step": 453
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 1.4776380400505558,
+      "learning_rate": 9.399000427690736e-06,
+      "loss": 1.0186,
+      "step": 454
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 1.4965281834004576,
+      "learning_rate": 9.394566789336707e-06,
+      "loss": 1.0291,
+      "step": 455
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 1.4653625188327892,
+      "learning_rate": 9.390117910963132e-06,
+      "loss": 1.0703,
+      "step": 456
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 1.462418393729156,
+      "learning_rate": 9.385653807998376e-06,
+      "loss": 0.9559,
+      "step": 457
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 1.3302119494978049,
+      "learning_rate": 9.381174495923608e-06,
+      "loss": 0.9573,
+      "step": 458
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 1.3552136652414049,
+      "learning_rate": 9.376679990272736e-06,
+      "loss": 1.0005,
+      "step": 459
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 1.2685383712045888,
+      "learning_rate": 9.37217030663236e-06,
+      "loss": 1.0049,
+      "step": 460
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 1.3936190052257311,
+      "learning_rate": 9.367645460641716e-06,
+      "loss": 0.9823,
+      "step": 461
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 1.4268026826598337,
+      "learning_rate": 9.36310546799262e-06,
+      "loss": 1.0155,
+      "step": 462
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 1.4533276720040302,
+      "learning_rate": 9.358550344429421e-06,
+      "loss": 1.0139,
+      "step": 463
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 1.521002659902884,
+      "learning_rate": 9.353980105748934e-06,
+      "loss": 0.8908,
+      "step": 464
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 1.641710919738433,
+      "learning_rate": 9.349394767800397e-06,
+      "loss": 1.0306,
+      "step": 465
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 1.4716078872980527,
+      "learning_rate": 9.344794346485408e-06,
+      "loss": 0.8803,
+      "step": 466
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 1.5122452528847365,
+      "learning_rate": 9.340178857757876e-06,
+      "loss": 0.9515,
+      "step": 467
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 1.5798633305945757,
+      "learning_rate": 9.335548317623957e-06,
+      "loss": 1.0285,
+      "step": 468
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 1.4307669033591586,
+      "learning_rate": 9.330902742142013e-06,
+      "loss": 0.9047,
+      "step": 469
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 1.8427803448268054,
+      "learning_rate": 9.326242147422538e-06,
+      "loss": 1.1001,
+      "step": 470
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 1.575779919679042,
+      "learning_rate": 9.321566549628118e-06,
+      "loss": 0.9324,
+      "step": 471
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 1.4123319976161952,
+      "learning_rate": 9.316875964973366e-06,
+      "loss": 1.0209,
+      "step": 472
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 1.6511900788251523,
+      "learning_rate": 9.31217040972487e-06,
+      "loss": 0.8342,
+      "step": 473
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 1.390908571577129,
+      "learning_rate": 9.307449900201132e-06,
+      "loss": 1.0583,
+      "step": 474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3872187347904474,
+      "learning_rate": 9.302714452772515e-06,
+      "loss": 0.9585,
+      "step": 475
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 1.450555938560158,
+      "learning_rate": 9.29796408386119e-06,
+      "loss": 1.052,
+      "step": 476
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 1.406397864368289,
+      "learning_rate": 9.293198809941067e-06,
+      "loss": 0.8413,
+      "step": 477
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 1.3931796985709013,
+      "learning_rate": 9.288418647537752e-06,
+      "loss": 0.7519,
+      "step": 478
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 1.5319054898122384,
+      "learning_rate": 9.283623613228479e-06,
+      "loss": 1.0298,
+      "step": 479
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 1.5432936514032687,
+      "learning_rate": 9.27881372364206e-06,
+      "loss": 0.9633,
+      "step": 480
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 1.5274086376229343,
+      "learning_rate": 9.27398899545882e-06,
+      "loss": 0.8164,
+      "step": 481
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 1.3980053759851052,
+      "learning_rate": 9.269149445410545e-06,
+      "loss": 0.9417,
+      "step": 482
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 1.49967062250857,
+      "learning_rate": 9.264295090280424e-06,
+      "loss": 0.9096,
+      "step": 483
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 1.4102222556674524,
+      "learning_rate": 9.259425946902987e-06,
+      "loss": 0.8849,
+      "step": 484
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 1.3685267843710929,
+      "learning_rate": 9.254542032164047e-06,
+      "loss": 0.9342,
+      "step": 485
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 1.4834858154279325,
+      "learning_rate": 9.249643363000645e-06,
+      "loss": 0.8741,
+      "step": 486
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 1.394239062821782,
+      "learning_rate": 9.24472995640099e-06,
+      "loss": 0.9932,
+      "step": 487
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 1.3564011350001746,
+      "learning_rate": 9.239801829404396e-06,
+      "loss": 0.8935,
+      "step": 488
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 1.2807001052427405,
+      "learning_rate": 9.234858999101232e-06,
+      "loss": 0.958,
+      "step": 489
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 1.4440503426953395,
+      "learning_rate": 9.22990148263285e-06,
+      "loss": 0.8924,
+      "step": 490
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 1.7351828274477326,
+      "learning_rate": 9.224929297191536e-06,
+      "loss": 1.0795,
+      "step": 491
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 1.4159219286177769,
+      "learning_rate": 9.219942460020447e-06,
+      "loss": 1.0378,
+      "step": 492
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 1.3241789957911565,
+      "learning_rate": 9.214940988413552e-06,
+      "loss": 0.936,
+      "step": 493
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 1.3581074529498713,
+      "learning_rate": 9.20992489971557e-06,
+      "loss": 0.8151,
+      "step": 494
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 1.770576835354788,
+      "learning_rate": 9.204894211321906e-06,
+      "loss": 0.8653,
+      "step": 495
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 1.4485875645863802,
+      "learning_rate": 9.199848940678607e-06,
+      "loss": 1.0535,
+      "step": 496
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 1.3898779265255505,
+      "learning_rate": 9.194789105282277e-06,
+      "loss": 0.958,
+      "step": 497
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 1.3461954362036133,
+      "learning_rate": 9.189714722680041e-06,
+      "loss": 0.8689,
+      "step": 498
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 1.4353523589481991,
+      "learning_rate": 9.184625810469468e-06,
+      "loss": 0.9563,
+      "step": 499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.238506915526249,
+      "learning_rate": 9.179522386298508e-06,
+      "loss": 0.9404,
+      "step": 500
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 1.370849780956035,
+      "learning_rate": 9.174404467865447e-06,
+      "loss": 0.9067,
+      "step": 501
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 1.4160629336418484,
+      "learning_rate": 9.169272072918834e-06,
+      "loss": 1.0436,
+      "step": 502
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 1.2951165265128666,
+      "learning_rate": 9.164125219257419e-06,
+      "loss": 0.9509,
+      "step": 503
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 1.2494437565297118,
+      "learning_rate": 9.158963924730092e-06,
+      "loss": 0.9967,
+      "step": 504
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 1.4623360028578793,
+      "learning_rate": 9.153788207235827e-06,
+      "loss": 0.9766,
+      "step": 505
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 1.3324285057386127,
+      "learning_rate": 9.148598084723615e-06,
+      "loss": 0.7784,
+      "step": 506
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 1.4749568020074846,
+      "learning_rate": 9.143393575192402e-06,
+      "loss": 0.9145,
+      "step": 507
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 1.3764109583409792,
+      "learning_rate": 9.138174696691025e-06,
+      "loss": 0.8893,
+      "step": 508
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 1.40301653728679,
+      "learning_rate": 9.132941467318152e-06,
+      "loss": 1.0235,
+      "step": 509
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 1.3912336384908608,
+      "learning_rate": 9.127693905222223e-06,
+      "loss": 0.9707,
+      "step": 510
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 1.4997379377361553,
+      "learning_rate": 9.122432028601377e-06,
+      "loss": 0.8917,
+      "step": 511
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 1.4322833624081515,
+      "learning_rate": 9.1171558557034e-06,
+      "loss": 1.038,
+      "step": 512
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 1.3929569631160912,
+      "learning_rate": 9.111865404825652e-06,
+      "loss": 0.8784,
+      "step": 513
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 1.637572058738915,
+      "learning_rate": 9.10656069431501e-06,
+      "loss": 0.8974,
+      "step": 514
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 1.2948647889172866,
+      "learning_rate": 9.101241742567802e-06,
+      "loss": 0.8767,
+      "step": 515
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 1.3408107051963223,
+      "learning_rate": 9.095908568029741e-06,
+      "loss": 0.8146,
+      "step": 516
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 1.474274545744325,
+      "learning_rate": 9.09056118919587e-06,
+      "loss": 0.8586,
+      "step": 517
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 1.3352798501271326,
+      "learning_rate": 9.085199624610486e-06,
+      "loss": 0.9243,
+      "step": 518
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 1.4717451332802163,
+      "learning_rate": 9.079823892867083e-06,
+      "loss": 0.8866,
+      "step": 519
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 1.5418753247936274,
+      "learning_rate": 9.074434012608282e-06,
+      "loss": 0.9183,
+      "step": 520
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 1.5053783608392823,
+      "learning_rate": 9.069030002525777e-06,
+      "loss": 1.0681,
+      "step": 521
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 1.3493309194564662,
+      "learning_rate": 9.063611881360258e-06,
+      "loss": 0.994,
+      "step": 522
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 1.5139291942099158,
+      "learning_rate": 9.05817966790135e-06,
+      "loss": 0.9318,
+      "step": 523
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 1.346865186676289,
+      "learning_rate": 9.052733380987555e-06,
+      "loss": 1.0488,
+      "step": 524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.411407629973773,
+      "learning_rate": 9.047273039506174e-06,
+      "loss": 0.8485,
+      "step": 525
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 1.3573160037669074,
+      "learning_rate": 9.041798662393255e-06,
+      "loss": 1.0622,
+      "step": 526
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 1.5045820525776132,
+      "learning_rate": 9.036310268633515e-06,
+      "loss": 0.9283,
+      "step": 527
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 1.6443751015760586,
+      "learning_rate": 9.030807877260278e-06,
+      "loss": 1.042,
+      "step": 528
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 1.4069340885537325,
+      "learning_rate": 9.025291507355419e-06,
+      "loss": 1.0123,
+      "step": 529
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 1.4468499985695986,
+      "learning_rate": 9.01976117804928e-06,
+      "loss": 1.0073,
+      "step": 530
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 1.491536292233139,
+      "learning_rate": 9.014216908520619e-06,
+      "loss": 1.0151,
+      "step": 531
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 1.4148302450423154,
+      "learning_rate": 9.008658717996538e-06,
+      "loss": 0.8005,
+      "step": 532
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 1.5002636294991516,
+      "learning_rate": 9.003086625752414e-06,
+      "loss": 0.8713,
+      "step": 533
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 1.4154094029725712,
+      "learning_rate": 8.997500651111833e-06,
+      "loss": 0.9189,
+      "step": 534
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 1.5940934634411033,
+      "learning_rate": 8.991900813446523e-06,
+      "loss": 1.1525,
+      "step": 535
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 1.4598501462944218,
+      "learning_rate": 8.986287132176295e-06,
+      "loss": 0.8486,
+      "step": 536
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 1.2206810298850677,
+      "learning_rate": 8.980659626768961e-06,
+      "loss": 0.8157,
+      "step": 537
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 1.366214691817922,
+      "learning_rate": 8.975018316740278e-06,
+      "loss": 0.8793,
+      "step": 538
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 1.4464236687534078,
+      "learning_rate": 8.969363221653875e-06,
+      "loss": 0.828,
+      "step": 539
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 1.504200571483841,
+      "learning_rate": 8.963694361121186e-06,
+      "loss": 0.889,
+      "step": 540
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 1.3741153204304815,
+      "learning_rate": 8.958011754801383e-06,
+      "loss": 0.8933,
+      "step": 541
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 1.5571331900681733,
+      "learning_rate": 8.952315422401307e-06,
+      "loss": 0.9756,
+      "step": 542
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 1.563803522445841,
+      "learning_rate": 8.946605383675403e-06,
+      "loss": 0.9677,
+      "step": 543
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 1.4223256865675364,
+      "learning_rate": 8.940881658425645e-06,
+      "loss": 0.7662,
+      "step": 544
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 1.5331985717149479,
+      "learning_rate": 8.93514426650147e-06,
+      "loss": 1.0973,
+      "step": 545
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 1.4592217601695612,
+      "learning_rate": 8.929393227799715e-06,
+      "loss": 0.9849,
+      "step": 546
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 1.35648945624887,
+      "learning_rate": 8.923628562264536e-06,
+      "loss": 1.0214,
+      "step": 547
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 1.3976902881668627,
+      "learning_rate": 8.917850289887353e-06,
+      "loss": 0.9981,
+      "step": 548
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 1.393326487055514,
+      "learning_rate": 8.91205843070677e-06,
+      "loss": 1.0041,
+      "step": 549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.622334625821984,
+      "learning_rate": 8.906253004808506e-06,
+      "loss": 0.9277,
+      "step": 550
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 1.3518556798628933,
+      "learning_rate": 8.900434032325332e-06,
+      "loss": 1.0108,
+      "step": 551
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 1.3453468202758352,
+      "learning_rate": 8.894601533437e-06,
+      "loss": 0.9626,
+      "step": 552
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 1.4142706573546748,
+      "learning_rate": 8.888755528370163e-06,
+      "loss": 0.8083,
+      "step": 553
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 1.4560792961358122,
+      "learning_rate": 8.882896037398322e-06,
+      "loss": 0.9766,
+      "step": 554
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 1.2912660906242206,
+      "learning_rate": 8.877023080841739e-06,
+      "loss": 0.8632,
+      "step": 555
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 1.282474100815103,
+      "learning_rate": 8.871136679067372e-06,
+      "loss": 0.8983,
+      "step": 556
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 1.3112839120354007,
+      "learning_rate": 8.865236852488813e-06,
+      "loss": 1.0322,
+      "step": 557
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 1.3621632259910548,
+      "learning_rate": 8.859323621566207e-06,
+      "loss": 1.0287,
+      "step": 558
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 1.5910374654894264,
+      "learning_rate": 8.853397006806183e-06,
+      "loss": 0.9828,
+      "step": 559
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 1.7360251741625654,
+      "learning_rate": 8.847457028761783e-06,
+      "loss": 0.9155,
+      "step": 560
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 1.3895171639579713,
+      "learning_rate": 8.841503708032398e-06,
+      "loss": 0.9202,
+      "step": 561
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 1.3975492240869014,
+      "learning_rate": 8.835537065263684e-06,
+      "loss": 0.9689,
+      "step": 562
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 1.5387715653908491,
+      "learning_rate": 8.829557121147499e-06,
+      "loss": 0.7787,
+      "step": 563
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 1.5403130406239296,
+      "learning_rate": 8.82356389642183e-06,
+      "loss": 0.8101,
+      "step": 564
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 1.4201712683956695,
+      "learning_rate": 8.817557411870717e-06,
+      "loss": 0.9049,
+      "step": 565
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 1.4311177509030442,
+      "learning_rate": 8.811537688324187e-06,
+      "loss": 1.1066,
+      "step": 566
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 1.6019997601518197,
+      "learning_rate": 8.805504746658183e-06,
+      "loss": 0.8932,
+      "step": 567
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 1.5346907398057228,
+      "learning_rate": 8.799458607794476e-06,
+      "loss": 0.9188,
+      "step": 568
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 1.3418191741239052,
+      "learning_rate": 8.793399292700616e-06,
+      "loss": 0.7528,
+      "step": 569
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 1.382352397042085,
+      "learning_rate": 8.787326822389836e-06,
+      "loss": 0.9525,
+      "step": 570
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 1.5669966399046997,
+      "learning_rate": 8.781241217921e-06,
+      "loss": 0.8999,
+      "step": 571
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 1.6401100746122428,
+      "learning_rate": 8.775142500398513e-06,
+      "loss": 0.8721,
+      "step": 572
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 1.358336625713025,
+      "learning_rate": 8.769030690972262e-06,
+      "loss": 0.9477,
+      "step": 573
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 1.7128352870499037,
+      "learning_rate": 8.76290581083753e-06,
+      "loss": 0.8941,
+      "step": 574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3994151039561413,
+      "learning_rate": 8.756767881234928e-06,
+      "loss": 1.0012,
+      "step": 575
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 1.4454731560472749,
+      "learning_rate": 8.750616923450328e-06,
+      "loss": 0.8908,
+      "step": 576
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 1.419632383768781,
+      "learning_rate": 8.744452958814775e-06,
+      "loss": 0.9682,
+      "step": 577
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 1.467796332389228,
+      "learning_rate": 8.738276008704426e-06,
+      "loss": 1.0111,
+      "step": 578
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 1.4604232569075957,
+      "learning_rate": 8.732086094540467e-06,
+      "loss": 0.924,
+      "step": 579
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 1.4717953084543596,
+      "learning_rate": 8.725883237789046e-06,
+      "loss": 0.9978,
+      "step": 580
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 1.4082099843914526,
+      "learning_rate": 8.719667459961191e-06,
+      "loss": 0.8686,
+      "step": 581
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 1.6474003041888512,
+      "learning_rate": 8.713438782612743e-06,
+      "loss": 0.9175,
+      "step": 582
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 1.3073897969502495,
+      "learning_rate": 8.707197227344275e-06,
+      "loss": 0.9628,
+      "step": 583
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 1.3152130346924975,
+      "learning_rate": 8.700942815801023e-06,
+      "loss": 0.9566,
+      "step": 584
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 1.3769807176772066,
+      "learning_rate": 8.6946755696728e-06,
+      "loss": 0.9675,
+      "step": 585
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 1.3923535033005527,
+      "learning_rate": 8.688395510693939e-06,
+      "loss": 0.8565,
+      "step": 586
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 1.4986518252287144,
+      "learning_rate": 8.682102660643196e-06,
+      "loss": 0.9421,
+      "step": 587
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 1.4722552033472776,
+      "learning_rate": 8.675797041343696e-06,
+      "loss": 0.9932,
+      "step": 588
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 1.3302095748418352,
+      "learning_rate": 8.669478674662839e-06,
+      "loss": 0.7199,
+      "step": 589
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 1.4560186060094533,
+      "learning_rate": 8.663147582512232e-06,
+      "loss": 0.8765,
+      "step": 590
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 1.5525527897094802,
+      "learning_rate": 8.65680378684762e-06,
+      "loss": 0.9584,
+      "step": 591
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 1.4910849428701267,
+      "learning_rate": 8.6504473096688e-06,
+      "loss": 0.9156,
+      "step": 592
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 1.4283774795561834,
+      "learning_rate": 8.64407817301954e-06,
+      "loss": 0.9388,
+      "step": 593
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 1.3981306558257078,
+      "learning_rate": 8.637696398987517e-06,
+      "loss": 0.866,
+      "step": 594
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 1.3805052801264885,
+      "learning_rate": 8.631302009704235e-06,
+      "loss": 0.8499,
+      "step": 595
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 1.3819047614321103,
+      "learning_rate": 8.624895027344943e-06,
+      "loss": 0.9925,
+      "step": 596
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 1.373858197863816,
+      "learning_rate": 8.618475474128563e-06,
+      "loss": 0.8219,
+      "step": 597
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 1.457164534293568,
+      "learning_rate": 8.61204337231761e-06,
+      "loss": 0.8931,
+      "step": 598
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 1.3633032487499086,
+      "learning_rate": 8.605598744218122e-06,
+      "loss": 0.9469,
+      "step": 599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4861248088551802,
+      "learning_rate": 8.599141612179572e-06,
+      "loss": 0.8376,
+      "step": 600
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 1.3955680245886384,
+      "learning_rate": 8.592671998594794e-06,
+      "loss": 0.7075,
+      "step": 601
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 1.369109178076816,
+      "learning_rate": 8.586189925899913e-06,
+      "loss": 1.0529,
+      "step": 602
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 1.3427712910637437,
+      "learning_rate": 8.57969541657426e-06,
+      "loss": 0.7588,
+      "step": 603
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 1.3608443355480078,
+      "learning_rate": 8.57318849314029e-06,
+      "loss": 0.8033,
+      "step": 604
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 1.7105298590531357,
+      "learning_rate": 8.566669178163513e-06,
+      "loss": 0.9374,
+      "step": 605
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 1.386360426394754,
+      "learning_rate": 8.560137494252416e-06,
+      "loss": 0.8014,
+      "step": 606
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 1.4510189857433264,
+      "learning_rate": 8.553593464058374e-06,
+      "loss": 0.9721,
+      "step": 607
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 1.340447114304886,
+      "learning_rate": 8.54703711027558e-06,
+      "loss": 0.7932,
+      "step": 608
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 1.4285335581683642,
+      "learning_rate": 8.540468455640964e-06,
+      "loss": 1.0418,
+      "step": 609
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 16.803896978761056,
+      "learning_rate": 8.533887522934114e-06,
+      "loss": 0.9427,
+      "step": 610
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 1.665721896368663,
+      "learning_rate": 8.527294334977201e-06,
+      "loss": 0.9956,
+      "step": 611
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 1.5171218006454494,
+      "learning_rate": 8.520688914634894e-06,
+      "loss": 1.0418,
+      "step": 612
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 1.470209467232657,
+      "learning_rate": 8.51407128481428e-06,
+      "loss": 0.8713,
+      "step": 613
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 1.346090684374857,
+      "learning_rate": 8.507441468464792e-06,
+      "loss": 0.8803,
+      "step": 614
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 1.458017891779872,
+      "learning_rate": 8.50079948857812e-06,
+      "loss": 0.7842,
+      "step": 615
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 1.4381766913280036,
+      "learning_rate": 8.494145368188143e-06,
+      "loss": 0.7045,
+      "step": 616
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 1.7829705770360416,
+      "learning_rate": 8.487479130370838e-06,
+      "loss": 0.8787,
+      "step": 617
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 1.446303937218383,
+      "learning_rate": 8.480800798244202e-06,
+      "loss": 0.9451,
+      "step": 618
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 1.4273277946413612,
+      "learning_rate": 8.47411039496818e-06,
+      "loss": 0.8296,
+      "step": 619
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 1.2764160651206857,
+      "learning_rate": 8.467407943744574e-06,
+      "loss": 0.8424,
+      "step": 620
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 1.4416231540108717,
+      "learning_rate": 8.460693467816972e-06,
+      "loss": 0.9079,
+      "step": 621
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 1.2384301031567844,
+      "learning_rate": 8.453966990470656e-06,
+      "loss": 0.7483,
+      "step": 622
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 1.5322024323708576,
+      "learning_rate": 8.447228535032536e-06,
+      "loss": 0.9079,
+      "step": 623
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 1.4765307800726852,
+      "learning_rate": 8.440478124871054e-06,
+      "loss": 0.9689,
+      "step": 624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4679727544921668,
+      "learning_rate": 8.433715783396115e-06,
+      "loss": 0.948,
+      "step": 625
+    },
+    {
+      "epoch": 1.0016,
+      "grad_norm": 1.3150463559924286,
+      "learning_rate": 8.426941534058999e-06,
+      "loss": 0.6379,
+      "step": 626
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 1.3382635596905816,
+      "learning_rate": 8.420155400352279e-06,
+      "loss": 0.6869,
+      "step": 627
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 1.3418711158443763,
+      "learning_rate": 8.413357405809748e-06,
+      "loss": 0.7021,
+      "step": 628
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 1.3244033776131163,
+      "learning_rate": 8.406547574006326e-06,
+      "loss": 0.6935,
+      "step": 629
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 1.3649807955427056,
+      "learning_rate": 8.399725928557985e-06,
+      "loss": 0.8212,
+      "step": 630
+    },
+    {
+      "epoch": 1.0096,
+      "grad_norm": 1.313022533618309,
+      "learning_rate": 8.39289249312167e-06,
+      "loss": 0.783,
+      "step": 631
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 1.3423988973648833,
+      "learning_rate": 8.386047291395208e-06,
+      "loss": 0.6337,
+      "step": 632
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 1.4447629126754589,
+      "learning_rate": 8.37919034711723e-06,
+      "loss": 0.6779,
+      "step": 633
+    },
+    {
+      "epoch": 1.0144,
+      "grad_norm": 1.3224164726574559,
+      "learning_rate": 8.372321684067092e-06,
+      "loss": 0.542,
+      "step": 634
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 1.5926696379191534,
+      "learning_rate": 8.36544132606479e-06,
+      "loss": 0.7795,
+      "step": 635
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 1.4456391681537013,
+      "learning_rate": 8.358549296970877e-06,
+      "loss": 0.3835,
+      "step": 636
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 1.1872760931188928,
+      "learning_rate": 8.351645620686377e-06,
+      "loss": 0.7666,
+      "step": 637
+    },
+    {
+      "epoch": 1.0208,
+      "grad_norm": 1.3357765808877722,
+      "learning_rate": 8.34473032115271e-06,
+      "loss": 0.8622,
+      "step": 638
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 1.2414919946425809,
+      "learning_rate": 8.337803422351602e-06,
+      "loss": 0.8632,
+      "step": 639
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 1.2659916609600506,
+      "learning_rate": 8.33086494830501e-06,
+      "loss": 0.4974,
+      "step": 640
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 1.25172661614131,
+      "learning_rate": 8.323914923075018e-06,
+      "loss": 0.7109,
+      "step": 641
+    },
+    {
+      "epoch": 1.0272,
+      "grad_norm": 1.2836363520186467,
+      "learning_rate": 8.316953370763788e-06,
+      "loss": 0.6157,
+      "step": 642
+    },
+    {
+      "epoch": 1.0288,
+      "grad_norm": 1.212075548477766,
+      "learning_rate": 8.309980315513444e-06,
+      "loss": 0.7195,
+      "step": 643
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 1.296158826719871,
+      "learning_rate": 8.302995781506007e-06,
+      "loss": 0.8782,
+      "step": 644
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 1.3946277865913659,
+      "learning_rate": 8.295999792963301e-06,
+      "loss": 0.7184,
+      "step": 645
+    },
+    {
+      "epoch": 1.0336,
+      "grad_norm": 1.3476564727612108,
+      "learning_rate": 8.288992374146878e-06,
+      "loss": 0.7435,
+      "step": 646
+    },
+    {
+      "epoch": 1.0352,
+      "grad_norm": 1.1865884893820302,
+      "learning_rate": 8.281973549357927e-06,
+      "loss": 0.8636,
+      "step": 647
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 1.1765589278115693,
+      "learning_rate": 8.274943342937191e-06,
+      "loss": 0.7396,
+      "step": 648
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 1.357579571656876,
+      "learning_rate": 8.267901779264889e-06,
+      "loss": 0.7293,
+      "step": 649
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4953418956329396,
+      "learning_rate": 8.260848882760616e-06,
+      "loss": 0.6515,
+      "step": 650
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 1.5065010131627834,
+      "learning_rate": 8.25378467788328e-06,
+      "loss": 0.9057,
+      "step": 651
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 1.4711358677236037,
+      "learning_rate": 8.246709189130997e-06,
+      "loss": 0.8104,
+      "step": 652
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 1.3347843545728548,
+      "learning_rate": 8.23962244104102e-06,
+      "loss": 0.7632,
+      "step": 653
+    },
+    {
+      "epoch": 1.0464,
+      "grad_norm": 1.2292072183462304,
+      "learning_rate": 8.232524458189644e-06,
+      "loss": 0.5603,
+      "step": 654
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 1.3907296050078812,
+      "learning_rate": 8.225415265192126e-06,
+      "loss": 0.6614,
+      "step": 655
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 1.4134509254802445,
+      "learning_rate": 8.218294886702606e-06,
+      "loss": 0.8461,
+      "step": 656
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 1.32815182293549,
+      "learning_rate": 8.211163347414005e-06,
+      "loss": 0.8152,
+      "step": 657
+    },
+    {
+      "epoch": 1.0528,
+      "grad_norm": 1.4486387482863485,
+      "learning_rate": 8.20402067205795e-06,
+      "loss": 0.7479,
+      "step": 658
+    },
+    {
+      "epoch": 1.0544,
+      "grad_norm": 1.265158500971521,
+      "learning_rate": 8.196866885404697e-06,
+      "loss": 0.7542,
+      "step": 659
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 1.3576648903080404,
+      "learning_rate": 8.18970201226302e-06,
+      "loss": 0.7166,
+      "step": 660
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 1.2946051626304147,
+      "learning_rate": 8.182526077480153e-06,
+      "loss": 0.8637,
+      "step": 661
+    },
+    {
+      "epoch": 1.0592,
+      "grad_norm": 1.573692891594583,
+      "learning_rate": 8.175339105941685e-06,
+      "loss": 0.5697,
+      "step": 662
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 1.3215955057696236,
+      "learning_rate": 8.168141122571478e-06,
+      "loss": 0.6819,
+      "step": 663
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 1.3495940845714642,
+      "learning_rate": 8.160932152331587e-06,
+      "loss": 0.7929,
+      "step": 664
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 1.2627390934138414,
+      "learning_rate": 8.153712220222163e-06,
+      "loss": 0.6596,
+      "step": 665
+    },
+    {
+      "epoch": 1.0656,
+      "grad_norm": 1.26329980422197,
+      "learning_rate": 8.14648135128138e-06,
+      "loss": 0.6535,
+      "step": 666
+    },
+    {
+      "epoch": 1.0672,
+      "grad_norm": 1.3103595918564983,
+      "learning_rate": 8.139239570585334e-06,
+      "loss": 0.6974,
+      "step": 667
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 1.2586847844323472,
+      "learning_rate": 8.131986903247959e-06,
+      "loss": 0.8504,
+      "step": 668
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 1.378025163477823,
+      "learning_rate": 8.124723374420951e-06,
+      "loss": 0.836,
+      "step": 669
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 1.2662107832079368,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.6784,
+      "step": 670
+    },
+    {
+      "epoch": 1.0735999999999999,
+      "grad_norm": 1.3512614263199396,
+      "learning_rate": 8.11016383309305e-06,
+      "loss": 0.686,
+      "step": 671
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 1.3858901555841285,
+      "learning_rate": 8.102867871083528e-06,
+      "loss": 0.8097,
+      "step": 672
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 1.2451277430024372,
+      "learning_rate": 8.095561148566932e-06,
+      "loss": 0.8129,
+      "step": 673
+    },
+    {
+      "epoch": 1.0784,
+      "grad_norm": 1.4480321809627075,
+      "learning_rate": 8.088243690882421e-06,
+      "loss": 0.6415,
+      "step": 674
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2906922302421495,
+      "learning_rate": 8.080915523406371e-06,
+      "loss": 0.7802,
+      "step": 675
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 1.2213875218649095,
+      "learning_rate": 8.073576671552303e-06,
+      "loss": 0.7211,
+      "step": 676
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 1.4313837644100027,
+      "learning_rate": 8.06622716077079e-06,
+      "loss": 0.6886,
+      "step": 677
+    },
+    {
+      "epoch": 1.0848,
+      "grad_norm": 1.1561446153650237,
+      "learning_rate": 8.058867016549372e-06,
+      "loss": 0.6541,
+      "step": 678
+    },
+    {
+      "epoch": 1.0864,
+      "grad_norm": 1.2373583788458533,
+      "learning_rate": 8.051496264412464e-06,
+      "loss": 0.7421,
+      "step": 679
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 1.3603153447773453,
+      "learning_rate": 8.044114929921264e-06,
+      "loss": 0.8531,
+      "step": 680
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 1.6058597890230129,
+      "learning_rate": 8.036723038673675e-06,
+      "loss": 0.8283,
+      "step": 681
+    },
+    {
+      "epoch": 1.0912,
+      "grad_norm": 1.247180934371467,
+      "learning_rate": 8.029320616304204e-06,
+      "loss": 0.7982,
+      "step": 682
+    },
+    {
+      "epoch": 1.0928,
+      "grad_norm": 1.472773560725052,
+      "learning_rate": 8.021907688483885e-06,
+      "loss": 0.7267,
+      "step": 683
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 1.4939804407968524,
+      "learning_rate": 8.01448428092018e-06,
+      "loss": 0.6559,
+      "step": 684
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 1.1731369665351246,
+      "learning_rate": 8.007050419356898e-06,
+      "loss": 0.7354,
+      "step": 685
+    },
+    {
+      "epoch": 1.0976,
+      "grad_norm": 1.2456127972073165,
+      "learning_rate": 7.999606129574096e-06,
+      "loss": 0.7986,
+      "step": 686
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 1.2078276419353555,
+      "learning_rate": 7.992151437387999e-06,
+      "loss": 0.7435,
+      "step": 687
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 1.1646996440572461,
+      "learning_rate": 7.984686368650907e-06,
+      "loss": 0.5385,
+      "step": 688
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 1.2578890318118718,
+      "learning_rate": 7.977210949251102e-06,
+      "loss": 0.6397,
+      "step": 689
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 1.268239674773961,
+      "learning_rate": 7.969725205112766e-06,
+      "loss": 0.6496,
+      "step": 690
+    },
+    {
+      "epoch": 1.1056,
+      "grad_norm": 1.2309838453108177,
+      "learning_rate": 7.962229162195882e-06,
+      "loss": 0.7191,
+      "step": 691
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 1.381815481108343,
+      "learning_rate": 7.95472284649615e-06,
+      "loss": 0.9031,
+      "step": 692
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 1.1874896159284005,
+      "learning_rate": 7.947206284044896e-06,
+      "loss": 0.5007,
+      "step": 693
+    },
+    {
+      "epoch": 1.1104,
+      "grad_norm": 1.376743327977063,
+      "learning_rate": 7.939679500908982e-06,
+      "loss": 0.8523,
+      "step": 694
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 1.3310161589886995,
+      "learning_rate": 7.932142523190711e-06,
+      "loss": 0.7465,
+      "step": 695
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 1.292131620690502,
+      "learning_rate": 7.924595377027741e-06,
+      "loss": 0.7991,
+      "step": 696
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 1.1893212488642204,
+      "learning_rate": 7.917038088592997e-06,
+      "loss": 0.7561,
+      "step": 697
+    },
+    {
+      "epoch": 1.1168,
+      "grad_norm": 1.2487283208322288,
+      "learning_rate": 7.90947068409457e-06,
+      "loss": 0.6213,
+      "step": 698
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 1.1155105238521892,
+      "learning_rate": 7.90189318977564e-06,
+      "loss": 0.6535,
+      "step": 699
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.2102414165698925,
+      "learning_rate": 7.894305631914373e-06,
+      "loss": 0.6265,
+      "step": 700
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 1.2753606634231442,
+      "learning_rate": 7.886708036823838e-06,
+      "loss": 0.8215,
+      "step": 701
+    },
+    {
+      "epoch": 1.1232,
+      "grad_norm": 1.203151093365828,
+      "learning_rate": 7.879100430851907e-06,
+      "loss": 0.5359,
+      "step": 702
+    },
+    {
+      "epoch": 1.1248,
+      "grad_norm": 1.2249878107111256,
+      "learning_rate": 7.871482840381174e-06,
+      "loss": 0.7969,
+      "step": 703
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 1.351634334885911,
+      "learning_rate": 7.863855291828857e-06,
+      "loss": 0.8814,
+      "step": 704
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 1.244554399045127,
+      "learning_rate": 7.856217811646707e-06,
+      "loss": 0.8609,
+      "step": 705
+    },
+    {
+      "epoch": 1.1296,
+      "grad_norm": 1.2544685995658411,
+      "learning_rate": 7.848570426320918e-06,
+      "loss": 0.6342,
+      "step": 706
+    },
+    {
+      "epoch": 1.1312,
+      "grad_norm": 1.271734960445676,
+      "learning_rate": 7.840913162372032e-06,
+      "loss": 0.6004,
+      "step": 707
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 1.2782013050756662,
+      "learning_rate": 7.833246046354856e-06,
+      "loss": 0.9004,
+      "step": 708
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 1.1678713343875637,
+      "learning_rate": 7.825569104858353e-06,
+      "loss": 0.3863,
+      "step": 709
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 1.1657228744704804,
+      "learning_rate": 7.81788236450557e-06,
+      "loss": 0.6311,
+      "step": 710
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 1.7329485500647337,
+      "learning_rate": 7.810185851953529e-06,
+      "loss": 0.6922,
+      "step": 711
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 1.4777413029123616,
+      "learning_rate": 7.802479593893142e-06,
+      "loss": 0.744,
+      "step": 712
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 1.1873162877543122,
+      "learning_rate": 7.794763617049124e-06,
+      "loss": 0.718,
+      "step": 713
+    },
+    {
+      "epoch": 1.1424,
+      "grad_norm": 1.4600035891890293,
+      "learning_rate": 7.787037948179884e-06,
+      "loss": 1.009,
+      "step": 714
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 1.347580042102973,
+      "learning_rate": 7.779302614077449e-06,
+      "loss": 0.8002,
+      "step": 715
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 1.2799418366804134,
+      "learning_rate": 7.771557641567363e-06,
+      "loss": 0.4749,
+      "step": 716
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 1.352482313570435,
+      "learning_rate": 7.763803057508594e-06,
+      "loss": 0.6942,
+      "step": 717
+    },
+    {
+      "epoch": 1.1488,
+      "grad_norm": 1.2484848706235903,
+      "learning_rate": 7.756038888793446e-06,
+      "loss": 0.5569,
+      "step": 718
+    },
+    {
+      "epoch": 1.1504,
+      "grad_norm": 1.1503542720076902,
+      "learning_rate": 7.748265162347455e-06,
+      "loss": 0.6619,
+      "step": 719
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 1.2226637125329667,
+      "learning_rate": 7.740481905129307e-06,
+      "loss": 0.7461,
+      "step": 720
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 1.22034340764546,
+      "learning_rate": 7.732689144130741e-06,
+      "loss": 0.7821,
+      "step": 721
+    },
+    {
+      "epoch": 1.1552,
+      "grad_norm": 1.329003052555343,
+      "learning_rate": 7.724886906376451e-06,
+      "loss": 0.7239,
+      "step": 722
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 1.589137450367587,
+      "learning_rate": 7.717075218923998e-06,
+      "loss": 0.5245,
+      "step": 723
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 1.3566791219293877,
+      "learning_rate": 7.709254108863714e-06,
+      "loss": 0.6752,
+      "step": 724
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2473161414110405,
+      "learning_rate": 7.701423603318605e-06,
+      "loss": 0.8026,
+      "step": 725
+    },
+    {
+      "epoch": 1.1616,
+      "grad_norm": 1.1428087665441862,
+      "learning_rate": 7.693583729444263e-06,
+      "loss": 0.6737,
+      "step": 726
+    },
+    {
+      "epoch": 1.1632,
+      "grad_norm": 1.2372789049890767,
+      "learning_rate": 7.685734514428767e-06,
+      "loss": 0.6457,
+      "step": 727
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 1.328551206345503,
+      "learning_rate": 7.677875985492591e-06,
+      "loss": 0.6787,
+      "step": 728
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 1.5073386185687028,
+      "learning_rate": 7.67000816988851e-06,
+      "loss": 0.6835,
+      "step": 729
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 1.320843576213995,
+      "learning_rate": 7.662131094901499e-06,
+      "loss": 0.7719,
+      "step": 730
+    },
+    {
+      "epoch": 1.1696,
+      "grad_norm": 1.225046816423029,
+      "learning_rate": 7.654244787848655e-06,
+      "loss": 0.7197,
+      "step": 731
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 1.2632129731715367,
+      "learning_rate": 7.646349276079079e-06,
+      "loss": 0.9252,
+      "step": 732
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 1.4646142783684537,
+      "learning_rate": 7.6384445869738e-06,
+      "loss": 0.8141,
+      "step": 733
+    },
+    {
+      "epoch": 1.1743999999999999,
+      "grad_norm": 1.229534753769994,
+      "learning_rate": 7.630530747945672e-06,
+      "loss": 0.5457,
+      "step": 734
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 1.1348283055607522,
+      "learning_rate": 7.622607786439279e-06,
+      "loss": 0.7019,
+      "step": 735
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 1.2448770104563307,
+      "learning_rate": 7.6146757299308406e-06,
+      "loss": 0.7208,
+      "step": 736
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 1.2430201979909319,
+      "learning_rate": 7.606734605928123e-06,
+      "loss": 0.5881,
+      "step": 737
+    },
+    {
+      "epoch": 1.1808,
+      "grad_norm": 1.2655569120365449,
+      "learning_rate": 7.598784441970329e-06,
+      "loss": 0.8044,
+      "step": 738
+    },
+    {
+      "epoch": 1.1824,
+      "grad_norm": 1.4317731854413938,
+      "learning_rate": 7.590825265628019e-06,
+      "loss": 0.8539,
+      "step": 739
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.176809986288581,
+      "learning_rate": 7.5828571045030005e-06,
+      "loss": 0.5923,
+      "step": 740
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 1.2896127987919543,
+      "learning_rate": 7.574879986228245e-06,
+      "loss": 0.8061,
+      "step": 741
+    },
+    {
+      "epoch": 1.1872,
+      "grad_norm": 1.206763374176362,
+      "learning_rate": 7.566893938467788e-06,
+      "loss": 0.7588,
+      "step": 742
+    },
+    {
+      "epoch": 1.1888,
+      "grad_norm": 1.2785428981842697,
+      "learning_rate": 7.558898988916624e-06,
+      "loss": 0.689,
+      "step": 743
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 1.4342390497373152,
+      "learning_rate": 7.550895165300626e-06,
+      "loss": 0.6614,
+      "step": 744
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 1.1723859512659018,
+      "learning_rate": 7.542882495376437e-06,
+      "loss": 0.5469,
+      "step": 745
+    },
+    {
+      "epoch": 1.1936,
+      "grad_norm": 1.3320005553893965,
+      "learning_rate": 7.5348610069313795e-06,
+      "loss": 0.8307,
+      "step": 746
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 1.247273520426172,
+      "learning_rate": 7.5268307277833605e-06,
+      "loss": 0.7256,
+      "step": 747
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 1.11016388309839,
+      "learning_rate": 7.518791685780769e-06,
+      "loss": 0.7602,
+      "step": 748
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 1.0972185339747849,
+      "learning_rate": 7.5107439088023845e-06,
+      "loss": 0.5134,
+      "step": 749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3147335196146237,
+      "learning_rate": 7.502687424757278e-06,
+      "loss": 0.7687,
+      "step": 750
+    },
+    {
+      "epoch": 1.2016,
+      "grad_norm": 1.1492368036729201,
+      "learning_rate": 7.4946222615847165e-06,
+      "loss": 0.7188,
+      "step": 751
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 1.1693322412172868,
+      "learning_rate": 7.486548447254065e-06,
+      "loss": 0.5955,
+      "step": 752
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 1.1961032842260078,
+      "learning_rate": 7.478466009764692e-06,
+      "loss": 0.7217,
+      "step": 753
+    },
+    {
+      "epoch": 1.2064,
+      "grad_norm": 1.2762156445465176,
+      "learning_rate": 7.470374977145867e-06,
+      "loss": 0.7836,
+      "step": 754
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 1.3973661580241465,
+      "learning_rate": 7.462275377456671e-06,
+      "loss": 0.6788,
+      "step": 755
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 1.2466155836988653,
+      "learning_rate": 7.4541672387858895e-06,
+      "loss": 0.5918,
+      "step": 756
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 1.2044569151566613,
+      "learning_rate": 7.446050589251928e-06,
+      "loss": 0.5956,
+      "step": 757
+    },
+    {
+      "epoch": 1.2128,
+      "grad_norm": 1.2447785625389778,
+      "learning_rate": 7.437925457002697e-06,
+      "loss": 0.7532,
+      "step": 758
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 1.264369711709651,
+      "learning_rate": 7.429791870215535e-06,
+      "loss": 0.4948,
+      "step": 759
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 1.7189292367398459,
+      "learning_rate": 7.421649857097092e-06,
+      "loss": 0.7372,
+      "step": 760
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 1.3134136796882032,
+      "learning_rate": 7.413499445883245e-06,
+      "loss": 0.7271,
+      "step": 761
+    },
+    {
+      "epoch": 1.2192,
+      "grad_norm": 1.2505174381107635,
+      "learning_rate": 7.405340664838994e-06,
+      "loss": 0.7695,
+      "step": 762
+    },
+    {
+      "epoch": 1.2208,
+      "grad_norm": 1.2813480088181017,
+      "learning_rate": 7.39717354225836e-06,
+      "loss": 0.8047,
+      "step": 763
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 1.1452393712205249,
+      "learning_rate": 7.3889981064643e-06,
+      "loss": 0.4978,
+      "step": 764
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 1.122607048884173,
+      "learning_rate": 7.380814385808594e-06,
+      "loss": 0.7725,
+      "step": 765
+    },
+    {
+      "epoch": 1.2256,
+      "grad_norm": 1.3107870462346884,
+      "learning_rate": 7.372622408671757e-06,
+      "loss": 0.6325,
+      "step": 766
+    },
+    {
+      "epoch": 1.2272,
+      "grad_norm": 1.2078186641214865,
+      "learning_rate": 7.364422203462935e-06,
+      "loss": 0.6594,
+      "step": 767
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 1.2684481861422428,
+      "learning_rate": 7.3562137986198065e-06,
+      "loss": 0.6647,
+      "step": 768
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 1.444134712296142,
+      "learning_rate": 7.3479972226084925e-06,
+      "loss": 0.8116,
+      "step": 769
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 1.3382684150436859,
+      "learning_rate": 7.339772503923445e-06,
+      "loss": 0.7905,
+      "step": 770
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 1.3744930430597597,
+      "learning_rate": 7.331539671087353e-06,
+      "loss": 0.5605,
+      "step": 771
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 1.2083922721443898,
+      "learning_rate": 7.32329875265105e-06,
+      "loss": 0.7185,
+      "step": 772
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 1.2008383142966526,
+      "learning_rate": 7.315049777193407e-06,
+      "loss": 0.7424,
+      "step": 773
+    },
+    {
+      "epoch": 1.2384,
+      "grad_norm": 1.1311883540219492,
+      "learning_rate": 7.306792773321234e-06,
+      "loss": 0.4657,
+      "step": 774
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1365107327962582,
+      "learning_rate": 7.298527769669188e-06,
+      "loss": 0.3332,
+      "step": 775
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 1.4315692593660396,
+      "learning_rate": 7.290254794899665e-06,
+      "loss": 0.7684,
+      "step": 776
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 1.1941790832183428,
+      "learning_rate": 7.281973877702705e-06,
+      "loss": 0.5097,
+      "step": 777
+    },
+    {
+      "epoch": 1.2448,
+      "grad_norm": 1.133046833901464,
+      "learning_rate": 7.2736850467958905e-06,
+      "loss": 0.6394,
+      "step": 778
+    },
+    {
+      "epoch": 1.2464,
+      "grad_norm": 1.313592874870108,
+      "learning_rate": 7.26538833092425e-06,
+      "loss": 0.5794,
+      "step": 779
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 1.2334977808904701,
+      "learning_rate": 7.257083758860159e-06,
+      "loss": 0.6137,
+      "step": 780
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 1.3059080466065442,
+      "learning_rate": 7.248771359403231e-06,
+      "loss": 0.7514,
+      "step": 781
+    },
+    {
+      "epoch": 1.2511999999999999,
+      "grad_norm": 1.2165203722285671,
+      "learning_rate": 7.240451161380226e-06,
+      "loss": 0.7221,
+      "step": 782
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 1.258527706263471,
+      "learning_rate": 7.232123193644957e-06,
+      "loss": 0.6155,
+      "step": 783
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 1.2450032302819374,
+      "learning_rate": 7.22378748507817e-06,
+      "loss": 0.7768,
+      "step": 784
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 1.3913480982521946,
+      "learning_rate": 7.215444064587462e-06,
+      "loss": 0.843,
+      "step": 785
+    },
+    {
+      "epoch": 1.2576,
+      "grad_norm": 1.1986392015647929,
+      "learning_rate": 7.207092961107176e-06,
+      "loss": 0.6213,
+      "step": 786
+    },
+    {
+      "epoch": 1.2591999999999999,
+      "grad_norm": 1.127053361773705,
+      "learning_rate": 7.198734203598294e-06,
+      "loss": 0.6848,
+      "step": 787
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 1.2984824292809738,
+      "learning_rate": 7.190367821048346e-06,
+      "loss": 0.5966,
+      "step": 788
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 1.3301178475419186,
+      "learning_rate": 7.181993842471301e-06,
+      "loss": 0.6409,
+      "step": 789
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 1.172097618757308,
+      "learning_rate": 7.173612296907473e-06,
+      "loss": 0.7531,
+      "step": 790
+    },
+    {
+      "epoch": 1.2656,
+      "grad_norm": 1.4614391378945681,
+      "learning_rate": 7.165223213423416e-06,
+      "loss": 0.594,
+      "step": 791
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 1.1122324457618662,
+      "learning_rate": 7.15682662111183e-06,
+      "loss": 0.5081,
+      "step": 792
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 1.464237332219084,
+      "learning_rate": 7.148422549091447e-06,
+      "loss": 0.9004,
+      "step": 793
+    },
+    {
+      "epoch": 1.2704,
+      "grad_norm": 1.3152098782323687,
+      "learning_rate": 7.140011026506945e-06,
+      "loss": 0.7216,
+      "step": 794
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 1.2202336437212022,
+      "learning_rate": 7.131592082528837e-06,
+      "loss": 0.6333,
+      "step": 795
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 1.444016159221979,
+      "learning_rate": 7.12316574635337e-06,
+      "loss": 0.6515,
+      "step": 796
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 1.360261649062722,
+      "learning_rate": 7.114732047202433e-06,
+      "loss": 0.7208,
+      "step": 797
+    },
+    {
+      "epoch": 1.2768,
+      "grad_norm": 1.2702057716599038,
+      "learning_rate": 7.106291014323445e-06,
+      "loss": 0.5431,
+      "step": 798
+    },
+    {
+      "epoch": 1.2784,
+      "grad_norm": 1.3703683579905175,
+      "learning_rate": 7.0978426769892585e-06,
+      "loss": 0.7329,
+      "step": 799
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2773188981175638,
+      "learning_rate": 7.089387064498057e-06,
+      "loss": 0.538,
+      "step": 800
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 1.2552660700161773,
+      "learning_rate": 7.080924206173253e-06,
+      "loss": 0.7355,
+      "step": 801
+    },
+    {
+      "epoch": 1.2832,
+      "grad_norm": 1.150631246649232,
+      "learning_rate": 7.072454131363391e-06,
+      "loss": 0.4949,
+      "step": 802
+    },
+    {
+      "epoch": 1.2848,
+      "grad_norm": 1.194319124754565,
+      "learning_rate": 7.063976869442037e-06,
+      "loss": 0.6453,
+      "step": 803
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 1.3559566272229402,
+      "learning_rate": 7.055492449807684e-06,
+      "loss": 0.8146,
+      "step": 804
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 1.3468025367859855,
+      "learning_rate": 7.047000901883646e-06,
+      "loss": 0.6787,
+      "step": 805
+    },
+    {
+      "epoch": 1.2896,
+      "grad_norm": 1.3136223209434104,
+      "learning_rate": 7.038502255117957e-06,
+      "loss": 0.6526,
+      "step": 806
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 1.1211789840758364,
+      "learning_rate": 7.029996538983273e-06,
+      "loss": 0.4298,
+      "step": 807
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 1.182543079647206,
+      "learning_rate": 7.021483782976759e-06,
+      "loss": 0.7771,
+      "step": 808
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 1.348464758071526,
+      "learning_rate": 7.012964016620002e-06,
+      "loss": 0.6979,
+      "step": 809
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 1.108665641723972,
+      "learning_rate": 7.004437269458894e-06,
+      "loss": 0.6426,
+      "step": 810
+    },
+    {
+      "epoch": 1.2976,
+      "grad_norm": 1.1958892845824423,
+      "learning_rate": 6.995903571063541e-06,
+      "loss": 0.763,
+      "step": 811
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 1.2447353087960016,
+      "learning_rate": 6.987362951028147e-06,
+      "loss": 0.6571,
+      "step": 812
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 1.1826537716209429,
+      "learning_rate": 6.97881543897093e-06,
+      "loss": 0.6902,
+      "step": 813
+    },
+    {
+      "epoch": 1.3024,
+      "grad_norm": 1.176845509701663,
+      "learning_rate": 6.970261064534003e-06,
+      "loss": 0.6404,
+      "step": 814
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 1.4108855214669127,
+      "learning_rate": 6.961699857383279e-06,
+      "loss": 0.6306,
+      "step": 815
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 46.53388245980635,
+      "learning_rate": 6.953131847208365e-06,
+      "loss": 1.0821,
+      "step": 816
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 1.2241873899022233,
+      "learning_rate": 6.944557063722459e-06,
+      "loss": 0.6337,
+      "step": 817
+    },
+    {
+      "epoch": 1.3088,
+      "grad_norm": 1.2946430111599534,
+      "learning_rate": 6.935975536662254e-06,
+      "loss": 0.6463,
+      "step": 818
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 1.2287500390738022,
+      "learning_rate": 6.9273872957878255e-06,
+      "loss": 0.752,
+      "step": 819
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 1.2257093956134881,
+      "learning_rate": 6.91879237088253e-06,
+      "loss": 0.7208,
+      "step": 820
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 1.304474490472767,
+      "learning_rate": 6.910190791752907e-06,
+      "loss": 0.55,
+      "step": 821
+    },
+    {
+      "epoch": 1.3152,
+      "grad_norm": 1.3529370515603025,
+      "learning_rate": 6.90158258822857e-06,
+      "loss": 0.5866,
+      "step": 822
+    },
+    {
+      "epoch": 1.3168,
+      "grad_norm": 1.1884484048174446,
+      "learning_rate": 6.892967790162109e-06,
+      "loss": 0.6201,
+      "step": 823
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 1.195995518863932,
+      "learning_rate": 6.884346427428978e-06,
+      "loss": 0.6539,
+      "step": 824
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.2908307744963217,
+      "learning_rate": 6.875718529927404e-06,
+      "loss": 0.6484,
+      "step": 825
+    },
+    {
+      "epoch": 1.3216,
+      "grad_norm": 1.2778832161250204,
+      "learning_rate": 6.867084127578267e-06,
+      "loss": 0.7368,
+      "step": 826
+    },
+    {
+      "epoch": 1.3232,
+      "grad_norm": 1.543008002660216,
+      "learning_rate": 6.858443250325013e-06,
+      "loss": 0.7004,
+      "step": 827
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 1.4165464938469805,
+      "learning_rate": 6.849795928133538e-06,
+      "loss": 0.6776,
+      "step": 828
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 1.394124666753603,
+      "learning_rate": 6.841142190992092e-06,
+      "loss": 0.8477,
+      "step": 829
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 1.1525270325247092,
+      "learning_rate": 6.832482068911167e-06,
+      "loss": 0.7357,
+      "step": 830
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 1.2516901993527954,
+      "learning_rate": 6.823815591923402e-06,
+      "loss": 0.7739,
+      "step": 831
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 1.3283591364444092,
+      "learning_rate": 6.815142790083473e-06,
+      "loss": 0.8599,
+      "step": 832
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 1.2150118416533005,
+      "learning_rate": 6.8064636934679885e-06,
+      "loss": 0.7176,
+      "step": 833
+    },
+    {
+      "epoch": 1.3344,
+      "grad_norm": 1.2949412968489789,
+      "learning_rate": 6.797778332175387e-06,
+      "loss": 0.8335,
+      "step": 834
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 1.2494961953813604,
+      "learning_rate": 6.789086736325834e-06,
+      "loss": 0.773,
+      "step": 835
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 1.197920461480709,
+      "learning_rate": 6.780388936061118e-06,
+      "loss": 0.845,
+      "step": 836
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 1.139479512470914,
+      "learning_rate": 6.771684961544537e-06,
+      "loss": 0.6957,
+      "step": 837
+    },
+    {
+      "epoch": 1.3408,
+      "grad_norm": 1.3339424134768902,
+      "learning_rate": 6.7629748429608076e-06,
+      "loss": 0.6768,
+      "step": 838
+    },
+    {
+      "epoch": 1.3424,
+      "grad_norm": 1.307659853669155,
+      "learning_rate": 6.754258610515949e-06,
+      "loss": 0.5684,
+      "step": 839
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 1.1961789114690817,
+      "learning_rate": 6.745536294437187e-06,
+      "loss": 0.7008,
+      "step": 840
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 1.1414621537606762,
+      "learning_rate": 6.736807924972841e-06,
+      "loss": 0.5492,
+      "step": 841
+    },
+    {
+      "epoch": 1.3472,
+      "grad_norm": 1.5457772776647454,
+      "learning_rate": 6.728073532392226e-06,
+      "loss": 0.7861,
+      "step": 842
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 1.260048162313854,
+      "learning_rate": 6.719333146985544e-06,
+      "loss": 0.6048,
+      "step": 843
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 1.3364964445698275,
+      "learning_rate": 6.710586799063777e-06,
+      "loss": 0.6599,
+      "step": 844
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 1.3454762763160635,
+      "learning_rate": 6.701834518958587e-06,
+      "loss": 0.7436,
+      "step": 845
+    },
+    {
+      "epoch": 1.3536000000000001,
+      "grad_norm": 1.2179239727757822,
+      "learning_rate": 6.6930763370222104e-06,
+      "loss": 0.6541,
+      "step": 846
+    },
+    {
+      "epoch": 1.3552,
+      "grad_norm": 1.1354658890083928,
+      "learning_rate": 6.684312283627348e-06,
+      "loss": 0.6203,
+      "step": 847
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 1.1967920518153194,
+      "learning_rate": 6.6755423891670605e-06,
+      "loss": 0.5918,
+      "step": 848
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 1.2733449968738881,
+      "learning_rate": 6.6667666840546685e-06,
+      "loss": 0.5086,
+      "step": 849
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 1.4476643016500503,
+      "learning_rate": 6.6579851987236435e-06,
+      "loss": 0.6687,
+      "step": 850
+    },
+    {
+      "epoch": 1.3616,
+      "grad_norm": 1.2133480658920532,
+      "learning_rate": 6.649197963627497e-06,
+      "loss": 0.6046,
+      "step": 851
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 1.224297480012605,
+      "learning_rate": 6.640405009239689e-06,
+      "loss": 0.6808,
+      "step": 852
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 1.046130541707376,
+      "learning_rate": 6.631606366053507e-06,
+      "loss": 0.4575,
+      "step": 853
+    },
+    {
+      "epoch": 1.3664,
+      "grad_norm": 1.171624614542443,
+      "learning_rate": 6.622802064581968e-06,
+      "loss": 0.6308,
+      "step": 854
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 1.3161781017870495,
+      "learning_rate": 6.613992135357713e-06,
+      "loss": 0.7714,
+      "step": 855
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 1.2110517802001237,
+      "learning_rate": 6.605176608932897e-06,
+      "loss": 0.6019,
+      "step": 856
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 1.2708060527866796,
+      "learning_rate": 6.596355515879091e-06,
+      "loss": 0.5552,
+      "step": 857
+    },
+    {
+      "epoch": 1.3728,
+      "grad_norm": 1.1187684959272344,
+      "learning_rate": 6.587528886787165e-06,
+      "loss": 0.6212,
+      "step": 858
+    },
+    {
+      "epoch": 1.3744,
+      "grad_norm": 1.2742922238672474,
+      "learning_rate": 6.578696752267189e-06,
+      "loss": 0.5689,
+      "step": 859
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 1.6963016413510086,
+      "learning_rate": 6.5698591429483286e-06,
+      "loss": 0.8115,
+      "step": 860
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 1.144929322088606,
+      "learning_rate": 6.5610160894787275e-06,
+      "loss": 0.5902,
+      "step": 861
+    },
+    {
+      "epoch": 1.3792,
+      "grad_norm": 1.1245879336213362,
+      "learning_rate": 6.552167622525421e-06,
+      "loss": 0.5814,
+      "step": 862
+    },
+    {
+      "epoch": 1.3808,
+      "grad_norm": 1.2701783221427057,
+      "learning_rate": 6.543313772774209e-06,
+      "loss": 0.6264,
+      "step": 863
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 1.248825009516559,
+      "learning_rate": 6.534454570929563e-06,
+      "loss": 0.6145,
+      "step": 864
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 1.325121841890245,
+      "learning_rate": 6.52559004771451e-06,
+      "loss": 0.7679,
+      "step": 865
+    },
+    {
+      "epoch": 1.3856,
+      "grad_norm": 1.205049854520623,
+      "learning_rate": 6.516720233870538e-06,
+      "loss": 0.5697,
+      "step": 866
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 1.2234793144063485,
+      "learning_rate": 6.507845160157476e-06,
+      "loss": 0.684,
+      "step": 867
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 1.2793767090423127,
+      "learning_rate": 6.498964857353401e-06,
+      "loss": 0.6191,
+      "step": 868
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 1.2450510848653478,
+      "learning_rate": 6.4900793562545165e-06,
+      "loss": 0.7763,
+      "step": 869
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 1.2300013854539367,
+      "learning_rate": 6.481188687675057e-06,
+      "loss": 0.5605,
+      "step": 870
+    },
+    {
+      "epoch": 1.3936,
+      "grad_norm": 1.0670244620385796,
+      "learning_rate": 6.47229288244718e-06,
+      "loss": 0.4697,
+      "step": 871
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 1.343794706775137,
+      "learning_rate": 6.46339197142085e-06,
+      "loss": 0.7798,
+      "step": 872
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 1.1756976713739082,
+      "learning_rate": 6.454485985463742e-06,
+      "loss": 0.7579,
+      "step": 873
+    },
+    {
+      "epoch": 1.3984,
+      "grad_norm": 1.3001663469437656,
+      "learning_rate": 6.445574955461134e-06,
+      "loss": 0.8896,
+      "step": 874
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3451160716452202,
+      "learning_rate": 6.436658912315789e-06,
+      "loss": 0.6161,
+      "step": 875
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 1.2456394893359914,
+      "learning_rate": 6.427737886947859e-06,
+      "loss": 0.6325,
+      "step": 876
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 1.3450311493157474,
+      "learning_rate": 6.418811910294776e-06,
+      "loss": 0.6892,
+      "step": 877
+    },
+    {
+      "epoch": 1.4048,
+      "grad_norm": 1.3464954533964244,
+      "learning_rate": 6.409881013311136e-06,
+      "loss": 0.8252,
+      "step": 878
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 1.1465914161129294,
+      "learning_rate": 6.400945226968607e-06,
+      "loss": 0.4747,
+      "step": 879
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 1.264273911093962,
+      "learning_rate": 6.392004582255807e-06,
+      "loss": 0.6252,
+      "step": 880
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 1.213733242964069,
+      "learning_rate": 6.383059110178205e-06,
+      "loss": 0.6381,
+      "step": 881
+    },
+    {
+      "epoch": 1.4112,
+      "grad_norm": 1.1468755403270334,
+      "learning_rate": 6.374108841758006e-06,
+      "loss": 0.7355,
+      "step": 882
+    },
+    {
+      "epoch": 1.4128,
+      "grad_norm": 1.1709436184179878,
+      "learning_rate": 6.365153808034057e-06,
+      "loss": 0.7031,
+      "step": 883
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 1.4408150622860891,
+      "learning_rate": 6.356194040061725e-06,
+      "loss": 0.8491,
+      "step": 884
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 1.0963033283520514,
+      "learning_rate": 6.3472295689127946e-06,
+      "loss": 0.5437,
+      "step": 885
+    },
+    {
+      "epoch": 1.4176,
+      "grad_norm": 1.21188504878786,
+      "learning_rate": 6.338260425675365e-06,
+      "loss": 0.6067,
+      "step": 886
+    },
+    {
+      "epoch": 1.4192,
+      "grad_norm": 1.1203230900896428,
+      "learning_rate": 6.329286641453729e-06,
+      "loss": 0.4556,
+      "step": 887
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 1.2464895366401267,
+      "learning_rate": 6.320308247368285e-06,
+      "loss": 0.6916,
+      "step": 888
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 1.2143239622749609,
+      "learning_rate": 6.311325274555413e-06,
+      "loss": 0.5412,
+      "step": 889
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 1.2535682013524978,
+      "learning_rate": 6.302337754167369e-06,
+      "loss": 0.4931,
+      "step": 890
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 1.2770871865968727,
+      "learning_rate": 6.2933457173721855e-06,
+      "loss": 0.6988,
+      "step": 891
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 1.0857766820747359,
+      "learning_rate": 6.2843491953535515e-06,
+      "loss": 0.5482,
+      "step": 892
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 1.0647522118160035,
+      "learning_rate": 6.275348219310715e-06,
+      "loss": 0.6045,
+      "step": 893
+    },
+    {
+      "epoch": 1.4304000000000001,
+      "grad_norm": 1.0113811764633014,
+      "learning_rate": 6.266342820458366e-06,
+      "loss": 0.4799,
+      "step": 894
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 1.330912544547094,
+      "learning_rate": 6.2573330300265375e-06,
+      "loss": 0.7209,
+      "step": 895
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 1.1688886937178378,
+      "learning_rate": 6.248318879260488e-06,
+      "loss": 0.609,
+      "step": 896
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 1.2596305948429973,
+      "learning_rate": 6.239300399420601e-06,
+      "loss": 0.7181,
+      "step": 897
+    },
+    {
+      "epoch": 1.4368,
+      "grad_norm": 1.32192027270104,
+      "learning_rate": 6.230277621782269e-06,
+      "loss": 0.7701,
+      "step": 898
+    },
+    {
+      "epoch": 1.4384000000000001,
+      "grad_norm": 1.4734853010334765,
+      "learning_rate": 6.221250577635791e-06,
+      "loss": 0.6579,
+      "step": 899
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3253100993077958,
+      "learning_rate": 6.2122192982862615e-06,
+      "loss": 0.758,
+      "step": 900
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 1.1372799791703023,
+      "learning_rate": 6.203183815053463e-06,
+      "loss": 0.5747,
+      "step": 901
+    },
+    {
+      "epoch": 1.4432,
+      "grad_norm": 1.0163785096496036,
+      "learning_rate": 6.1941441592717564e-06,
+      "loss": 0.719,
+      "step": 902
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 1.182850292246135,
+      "learning_rate": 6.185100362289972e-06,
+      "loss": 0.7374,
+      "step": 903
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 1.1513584088837014,
+      "learning_rate": 6.176052455471302e-06,
+      "loss": 0.5785,
+      "step": 904
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 1.4712041458432799,
+      "learning_rate": 6.167000470193189e-06,
+      "loss": 0.6012,
+      "step": 905
+    },
+    {
+      "epoch": 1.4496,
+      "grad_norm": 1.4314422197677397,
+      "learning_rate": 6.157944437847226e-06,
+      "loss": 0.7326,
+      "step": 906
+    },
+    {
+      "epoch": 1.4512,
+      "grad_norm": 1.1170620992240652,
+      "learning_rate": 6.148884389839035e-06,
+      "loss": 0.5299,
+      "step": 907
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 1.1229635286673914,
+      "learning_rate": 6.1398203575881645e-06,
+      "loss": 0.6806,
+      "step": 908
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 1.3091673234744405,
+      "learning_rate": 6.130752372527981e-06,
+      "loss": 0.5943,
+      "step": 909
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.2085663488823692,
+      "learning_rate": 6.121680466105559e-06,
+      "loss": 0.7279,
+      "step": 910
+    },
+    {
+      "epoch": 1.4576,
+      "grad_norm": 1.0751539649030017,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.6444,
+      "step": 911
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 1.2494570171094106,
+      "learning_rate": 6.1035250150301864e-06,
+      "loss": 0.7919,
+      "step": 912
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 1.2489476958452004,
+      "learning_rate": 6.0944415333389405e-06,
+      "loss": 0.7331,
+      "step": 913
+    },
+    {
+      "epoch": 1.4624,
+      "grad_norm": 1.2106767017312203,
+      "learning_rate": 6.085354256208655e-06,
+      "loss": 0.7848,
+      "step": 914
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 1.170426818322747,
+      "learning_rate": 6.076263215153308e-06,
+      "loss": 0.7152,
+      "step": 915
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 1.1080045204919717,
+      "learning_rate": 6.067168441699927e-06,
+      "loss": 0.561,
+      "step": 916
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 1.1634797821208192,
+      "learning_rate": 6.058069967388489e-06,
+      "loss": 0.6304,
+      "step": 917
+    },
+    {
+      "epoch": 1.4687999999999999,
+      "grad_norm": 1.0365563365268824,
+      "learning_rate": 6.048967823771802e-06,
+      "loss": 0.6906,
+      "step": 918
+    },
+    {
+      "epoch": 1.4704,
+      "grad_norm": 1.1237783232674887,
+      "learning_rate": 6.039862042415401e-06,
+      "loss": 0.7177,
+      "step": 919
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 1.2690238204302238,
+      "learning_rate": 6.030752654897435e-06,
+      "loss": 0.7318,
+      "step": 920
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 1.2420366244925762,
+      "learning_rate": 6.021639692808558e-06,
+      "loss": 0.6329,
+      "step": 921
+    },
+    {
+      "epoch": 1.4752,
+      "grad_norm": 1.1385998159982653,
+      "learning_rate": 6.0125231877518205e-06,
+      "loss": 0.8191,
+      "step": 922
+    },
+    {
+      "epoch": 1.4768,
+      "grad_norm": 1.257650592109921,
+      "learning_rate": 6.0034031713425636e-06,
+      "loss": 0.7229,
+      "step": 923
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 1.3363318654253598,
+      "learning_rate": 5.994279675208302e-06,
+      "loss": 0.6503,
+      "step": 924
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1379796817349264,
+      "learning_rate": 5.985152730988617e-06,
+      "loss": 0.6202,
+      "step": 925
+    },
+    {
+      "epoch": 1.4816,
+      "grad_norm": 0.9842652996970314,
+      "learning_rate": 5.9760223703350495e-06,
+      "loss": 0.4728,
+      "step": 926
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 1.085294597562177,
+      "learning_rate": 5.966888624910989e-06,
+      "loss": 0.4007,
+      "step": 927
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 1.2196385559914007,
+      "learning_rate": 5.957751526391558e-06,
+      "loss": 0.6416,
+      "step": 928
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 1.1253266935556596,
+      "learning_rate": 5.948611106463518e-06,
+      "loss": 0.7154,
+      "step": 929
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 1.2253562369505184,
+      "learning_rate": 5.939467396825137e-06,
+      "loss": 0.666,
+      "step": 930
+    },
+    {
+      "epoch": 1.4896,
+      "grad_norm": 1.1183374053307045,
+      "learning_rate": 5.9303204291860975e-06,
+      "loss": 0.4879,
+      "step": 931
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 1.1586755535219952,
+      "learning_rate": 5.92117023526738e-06,
+      "loss": 0.7857,
+      "step": 932
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 1.3131283523378914,
+      "learning_rate": 5.912016846801153e-06,
+      "loss": 0.7079,
+      "step": 933
+    },
+    {
+      "epoch": 1.4944,
+      "grad_norm": 1.0980315908994558,
+      "learning_rate": 5.902860295530665e-06,
+      "loss": 0.7171,
+      "step": 934
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 0.9394132965471043,
+      "learning_rate": 5.893700613210128e-06,
+      "loss": 0.4202,
+      "step": 935
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 1.2058695978401652,
+      "learning_rate": 5.88453783160462e-06,
+      "loss": 0.5488,
+      "step": 936
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 1.0793124728891805,
+      "learning_rate": 5.875371982489959e-06,
+      "loss": 0.3475,
+      "step": 937
+    },
+    {
+      "epoch": 1.5008,
+      "grad_norm": 1.328873025037291,
+      "learning_rate": 5.866203097652605e-06,
+      "loss": 0.6553,
+      "step": 938
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 1.333670569224969,
+      "learning_rate": 5.857031208889548e-06,
+      "loss": 0.8131,
+      "step": 939
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 1.3407004855688325,
+      "learning_rate": 5.847856348008188e-06,
+      "loss": 0.7184,
+      "step": 940
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 1.11308873725126,
+      "learning_rate": 5.838678546826242e-06,
+      "loss": 0.5408,
+      "step": 941
+    },
+    {
+      "epoch": 1.5072,
+      "grad_norm": 1.1833722341674666,
+      "learning_rate": 5.829497837171616e-06,
+      "loss": 0.6573,
+      "step": 942
+    },
+    {
+      "epoch": 1.5088,
+      "grad_norm": 1.256678635823589,
+      "learning_rate": 5.820314250882304e-06,
+      "loss": 0.6949,
+      "step": 943
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 1.0971453330659688,
+      "learning_rate": 5.811127819806277e-06,
+      "loss": 0.4679,
+      "step": 944
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 1.0378575897456928,
+      "learning_rate": 5.801938575801372e-06,
+      "loss": 0.6328,
+      "step": 945
+    },
+    {
+      "epoch": 1.5135999999999998,
+      "grad_norm": 1.0350918399640188,
+      "learning_rate": 5.792746550735182e-06,
+      "loss": 0.5345,
+      "step": 946
+    },
+    {
+      "epoch": 1.5152,
+      "grad_norm": 1.0665133628136902,
+      "learning_rate": 5.7835517764849395e-06,
+      "loss": 0.6884,
+      "step": 947
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 1.0993398456814216,
+      "learning_rate": 5.7743542849374155e-06,
+      "loss": 0.4466,
+      "step": 948
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 1.0927270488899437,
+      "learning_rate": 5.765154107988803e-06,
+      "loss": 0.7235,
+      "step": 949
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2788033367468394,
+      "learning_rate": 5.755951277544607e-06,
+      "loss": 0.905,
+      "step": 950
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 1.2071339841025914,
+      "learning_rate": 5.746745825519539e-06,
+      "loss": 0.7306,
+      "step": 951
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 1.1487977920286807,
+      "learning_rate": 5.737537783837395e-06,
+      "loss": 0.7155,
+      "step": 952
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 1.2748184523140151,
+      "learning_rate": 5.728327184430955e-06,
+      "loss": 0.5605,
+      "step": 953
+    },
+    {
+      "epoch": 1.5264,
+      "grad_norm": 1.2201830121827055,
+      "learning_rate": 5.719114059241871e-06,
+      "loss": 0.5578,
+      "step": 954
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 1.3675681106619149,
+      "learning_rate": 5.709898440220552e-06,
+      "loss": 0.822,
+      "step": 955
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 1.189175983240062,
+      "learning_rate": 5.700680359326055e-06,
+      "loss": 0.8151,
+      "step": 956
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 1.1767360888149707,
+      "learning_rate": 5.691459848525977e-06,
+      "loss": 0.723,
+      "step": 957
+    },
+    {
+      "epoch": 1.5328,
+      "grad_norm": 1.1271530436612984,
+      "learning_rate": 5.682236939796337e-06,
+      "loss": 0.5896,
+      "step": 958
+    },
+    {
+      "epoch": 1.5344,
+      "grad_norm": 1.135734163295131,
+      "learning_rate": 5.673011665121477e-06,
+      "loss": 0.5928,
+      "step": 959
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 1.201137374939531,
+      "learning_rate": 5.663784056493936e-06,
+      "loss": 0.6753,
+      "step": 960
+    },
+    {
+      "epoch": 1.5375999999999999,
+      "grad_norm": 1.0212172069410483,
+      "learning_rate": 5.6545541459143535e-06,
+      "loss": 0.4069,
+      "step": 961
+    },
+    {
+      "epoch": 1.5392000000000001,
+      "grad_norm": 1.0130155617464967,
+      "learning_rate": 5.6453219653913495e-06,
+      "loss": 0.6102,
+      "step": 962
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 1.1054964976598196,
+      "learning_rate": 5.636087546941413e-06,
+      "loss": 0.6462,
+      "step": 963
+    },
+    {
+      "epoch": 1.5424,
+      "grad_norm": 1.1133110056866564,
+      "learning_rate": 5.6268509225888005e-06,
+      "loss": 0.7431,
+      "step": 964
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 1.0359492378228154,
+      "learning_rate": 5.617612124365411e-06,
+      "loss": 0.6652,
+      "step": 965
+    },
+    {
+      "epoch": 1.5455999999999999,
+      "grad_norm": 1.1326266587880558,
+      "learning_rate": 5.608371184310688e-06,
+      "loss": 0.5085,
+      "step": 966
+    },
+    {
+      "epoch": 1.5472000000000001,
+      "grad_norm": 1.1720167595406872,
+      "learning_rate": 5.5991281344714984e-06,
+      "loss": 0.5763,
+      "step": 967
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 1.078688326869843,
+      "learning_rate": 5.5898830069020325e-06,
+      "loss": 0.697,
+      "step": 968
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 1.2385480892848886,
+      "learning_rate": 5.580635833663679e-06,
+      "loss": 0.5429,
+      "step": 969
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 1.3563383296219322,
+      "learning_rate": 5.5713866468249235e-06,
+      "loss": 0.6982,
+      "step": 970
+    },
+    {
+      "epoch": 1.5535999999999999,
+      "grad_norm": 1.1605383986560867,
+      "learning_rate": 5.562135478461234e-06,
+      "loss": 0.6992,
+      "step": 971
+    },
+    {
+      "epoch": 1.5552000000000001,
+      "grad_norm": 1.0220305596940367,
+      "learning_rate": 5.55288236065495e-06,
+      "loss": 0.594,
+      "step": 972
+    },
+    {
+      "epoch": 1.5568,
+      "grad_norm": 1.1919100406435972,
+      "learning_rate": 5.5436273254951734e-06,
+      "loss": 0.6217,
+      "step": 973
+    },
+    {
+      "epoch": 1.5584,
+      "grad_norm": 1.230831289548136,
+      "learning_rate": 5.5343704050776535e-06,
+      "loss": 0.8309,
+      "step": 974
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2788006090792805,
+      "learning_rate": 5.5251116315046785e-06,
+      "loss": 0.7837,
+      "step": 975
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 1.0787247343229602,
+      "learning_rate": 5.515851036884964e-06,
+      "loss": 0.6083,
+      "step": 976
+    },
+    {
+      "epoch": 1.5632000000000001,
+      "grad_norm": 1.259166774826267,
+      "learning_rate": 5.5065886533335355e-06,
+      "loss": 0.7687,
+      "step": 977
+    },
+    {
+      "epoch": 1.5648,
+      "grad_norm": 1.2117713158437649,
+      "learning_rate": 5.497324512971632e-06,
+      "loss": 0.5497,
+      "step": 978
+    },
+    {
+      "epoch": 1.5664,
+      "grad_norm": 1.158179895660278,
+      "learning_rate": 5.4880586479265774e-06,
+      "loss": 0.749,
+      "step": 979
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 1.1727442168388174,
+      "learning_rate": 5.478791090331677e-06,
+      "loss": 0.8196,
+      "step": 980
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 1.1657506382845677,
+      "learning_rate": 5.4695218723261115e-06,
+      "loss": 0.6108,
+      "step": 981
+    },
+    {
+      "epoch": 1.5712000000000002,
+      "grad_norm": 1.2842367514350705,
+      "learning_rate": 5.46025102605481e-06,
+      "loss": 0.8495,
+      "step": 982
+    },
+    {
+      "epoch": 1.5728,
+      "grad_norm": 1.297843482398709,
+      "learning_rate": 5.4509785836683606e-06,
+      "loss": 0.8309,
+      "step": 983
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 1.337332312874649,
+      "learning_rate": 5.441704577322877e-06,
+      "loss": 0.6943,
+      "step": 984
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 1.2481229888856822,
+      "learning_rate": 5.4324290391798995e-06,
+      "loss": 0.7741,
+      "step": 985
+    },
+    {
+      "epoch": 1.5776,
+      "grad_norm": 1.0785239140499205,
+      "learning_rate": 5.423152001406282e-06,
+      "loss": 0.6596,
+      "step": 986
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 0.9698373227693667,
+      "learning_rate": 5.413873496174077e-06,
+      "loss": 0.3542,
+      "step": 987
+    },
+    {
+      "epoch": 1.5808,
+      "grad_norm": 1.3507374786568112,
+      "learning_rate": 5.404593555660424e-06,
+      "loss": 0.5497,
+      "step": 988
+    },
+    {
+      "epoch": 1.5824,
+      "grad_norm": 1.110645766414615,
+      "learning_rate": 5.39531221204745e-06,
+      "loss": 0.8019,
+      "step": 989
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 1.1503283675821023,
+      "learning_rate": 5.3860294975221335e-06,
+      "loss": 0.683,
+      "step": 990
+    },
+    {
+      "epoch": 1.5856,
+      "grad_norm": 1.0824947964858467,
+      "learning_rate": 5.376745444276219e-06,
+      "loss": 0.583,
+      "step": 991
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 1.10801137275184,
+      "learning_rate": 5.3674600845060856e-06,
+      "loss": 0.7615,
+      "step": 992
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 1.0538255963787007,
+      "learning_rate": 5.358173450412649e-06,
+      "loss": 0.6623,
+      "step": 993
+    },
+    {
+      "epoch": 1.5904,
+      "grad_norm": 1.125773155438969,
+      "learning_rate": 5.34888557420124e-06,
+      "loss": 0.3727,
+      "step": 994
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 1.1948547915824514,
+      "learning_rate": 5.339596488081501e-06,
+      "loss": 0.5956,
+      "step": 995
+    },
+    {
+      "epoch": 1.5936,
+      "grad_norm": 1.2178388697334486,
+      "learning_rate": 5.330306224267268e-06,
+      "loss": 0.6146,
+      "step": 996
+    },
+    {
+      "epoch": 1.5952,
+      "grad_norm": 1.1861357676103677,
+      "learning_rate": 5.321014814976459e-06,
+      "loss": 0.7278,
+      "step": 997
+    },
+    {
+      "epoch": 1.5968,
+      "grad_norm": 1.121405836843168,
+      "learning_rate": 5.311722292430966e-06,
+      "loss": 0.6748,
+      "step": 998
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 1.0966607847446037,
+      "learning_rate": 5.302428688856544e-06,
+      "loss": 0.7073,
+      "step": 999
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9909081884464844,
+      "learning_rate": 5.293134036482697e-06,
+      "loss": 0.538,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6016,
+      "grad_norm": 1.211644834472217,
+      "learning_rate": 5.283838367542562e-06,
+      "loss": 0.6077,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6032,
+      "grad_norm": 1.1355196371419298,
+      "learning_rate": 5.274541714272805e-06,
+      "loss": 0.5972,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6048,
+      "grad_norm": 1.3598158401383813,
+      "learning_rate": 5.265244108913503e-06,
+      "loss": 0.7216,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6064,
+      "grad_norm": 1.2050225332270803,
+      "learning_rate": 5.255945583708037e-06,
+      "loss": 0.6453,
+      "step": 1004
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 1.0480979665816974,
+      "learning_rate": 5.2466461709029755e-06,
+      "loss": 0.6065,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6096,
+      "grad_norm": 1.1374243626816265,
+      "learning_rate": 5.237345902747969e-06,
+      "loss": 0.6764,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6112,
+      "grad_norm": 1.0409066453451836,
+      "learning_rate": 5.228044811495632e-06,
+      "loss": 0.6906,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 1.1517380868918468,
+      "learning_rate": 5.218742929401432e-06,
+      "loss": 0.7163,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6143999999999998,
+      "grad_norm": 1.0627718889959366,
+      "learning_rate": 5.2094402887235805e-06,
+      "loss": 0.6553,
+      "step": 1009
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 1.0696188488145764,
+      "learning_rate": 5.200136921722919e-06,
+      "loss": 0.5693,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 0.9814968568753212,
+      "learning_rate": 5.1908328606628114e-06,
+      "loss": 0.4051,
+      "step": 1011
+    },
+    {
+      "epoch": 1.6192,
+      "grad_norm": 1.1370639387574801,
+      "learning_rate": 5.181528137809023e-06,
+      "loss": 0.6475,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6208,
+      "grad_norm": 1.1375876132430442,
+      "learning_rate": 5.1722227854296195e-06,
+      "loss": 0.6229,
+      "step": 1013
+    },
+    {
+      "epoch": 1.6223999999999998,
+      "grad_norm": 1.001399029145848,
+      "learning_rate": 5.162916835794843e-06,
+      "loss": 0.4343,
+      "step": 1014
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 1.2210922815913183,
+      "learning_rate": 5.1536103211770135e-06,
+      "loss": 0.8521,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 1.1474632749193912,
+      "learning_rate": 5.14430327385041e-06,
+      "loss": 0.7631,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 1.1971639744728668,
+      "learning_rate": 5.134995726091152e-06,
+      "loss": 0.474,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6288,
+      "grad_norm": 1.3314526294409563,
+      "learning_rate": 5.1256877101771015e-06,
+      "loss": 0.4618,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6303999999999998,
+      "grad_norm": 1.1210823495743871,
+      "learning_rate": 5.116379258387742e-06,
+      "loss": 0.7,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 1.2291112808003997,
+      "learning_rate": 5.1070704030040675e-06,
+      "loss": 0.77,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6336,
+      "grad_norm": 1.2416530477644234,
+      "learning_rate": 5.097761176308471e-06,
+      "loss": 0.7892,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6352,
+      "grad_norm": 1.3725108883015724,
+      "learning_rate": 5.088451610584638e-06,
+      "loss": 0.716,
+      "step": 1022
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 1.1674581377913043,
+      "learning_rate": 5.079141738117423e-06,
+      "loss": 0.7369,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 1.235865331948493,
+      "learning_rate": 5.06983159119275e-06,
+      "loss": 0.7201,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 1.1766425486549785,
+      "learning_rate": 5.060521202097491e-06,
+      "loss": 0.6143,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6416,
+      "grad_norm": 1.1181639469338849,
+      "learning_rate": 5.051210603119358e-06,
+      "loss": 0.7181,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6432,
+      "grad_norm": 1.1763984392105327,
+      "learning_rate": 5.041899826546791e-06,
+      "loss": 0.5734,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6448,
+      "grad_norm": 0.886731758445484,
+      "learning_rate": 5.032588904668851e-06,
+      "loss": 0.5226,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 1.358183730897876,
+      "learning_rate": 5.023277869775097e-06,
+      "loss": 0.8088,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 1.0761040548279244,
+      "learning_rate": 5.013966754155482e-06,
+      "loss": 0.6791,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6496,
+      "grad_norm": 1.150528259121588,
+      "learning_rate": 5.004655590100238e-06,
+      "loss": 0.7183,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 1.0559296584200513,
+      "learning_rate": 4.995344409899764e-06,
+      "loss": 0.6048,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6528,
+      "grad_norm": 1.0886940103108016,
+      "learning_rate": 4.986033245844519e-06,
+      "loss": 0.6635,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6543999999999999,
+      "grad_norm": 1.266678172480723,
+      "learning_rate": 4.976722130224904e-06,
+      "loss": 0.8285,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 1.102110861134063,
+      "learning_rate": 4.967411095331149e-06,
+      "loss": 0.6563,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6576,
+      "grad_norm": 1.1148389787360238,
+      "learning_rate": 4.95810017345321e-06,
+      "loss": 0.5594,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6592,
+      "grad_norm": 1.1580113627302566,
+      "learning_rate": 4.948789396880644e-06,
+      "loss": 0.4618,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6608,
+      "grad_norm": 0.9736790482670147,
+      "learning_rate": 4.939478797902512e-06,
+      "loss": 0.5523,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6623999999999999,
+      "grad_norm": 1.1598292183989116,
+      "learning_rate": 4.930168408807252e-06,
+      "loss": 0.7533,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 1.1387778597664762,
+      "learning_rate": 4.920858261882578e-06,
+      "loss": 0.582,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 1.1115978786386556,
+      "learning_rate": 4.911548389415363e-06,
+      "loss": 0.5688,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6672,
+      "grad_norm": 0.9997337398578393,
+      "learning_rate": 4.9022388236915306e-06,
+      "loss": 0.6925,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6688,
+      "grad_norm": 1.2049238065331527,
+      "learning_rate": 4.892929596995934e-06,
+      "loss": 0.8847,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6703999999999999,
+      "grad_norm": 1.1715510037458747,
+      "learning_rate": 4.883620741612259e-06,
+      "loss": 0.7546,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 1.0681399831300908,
+      "learning_rate": 4.8743122898229e-06,
+      "loss": 0.6568,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6736,
+      "grad_norm": 1.1276248067659158,
+      "learning_rate": 4.865004273908851e-06,
+      "loss": 0.5168,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 1.3305906510955694,
+      "learning_rate": 4.855696726149593e-06,
+      "loss": 0.737,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 1.1134600215136596,
+      "learning_rate": 4.846389678822987e-06,
+      "loss": 0.6535,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6784,
+      "grad_norm": 1.406950318765235,
+      "learning_rate": 4.837083164205159e-06,
+      "loss": 0.7967,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 1.0788514474381512,
+      "learning_rate": 4.827777214570384e-06,
+      "loss": 0.5824,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6816,
+      "grad_norm": 1.169014447171907,
+      "learning_rate": 4.818471862190979e-06,
+      "loss": 0.7004,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6832,
+      "grad_norm": 1.2644365762660572,
+      "learning_rate": 4.809167139337191e-06,
+      "loss": 0.6356,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 1.415610108414829,
+      "learning_rate": 4.799863078277082e-06,
+      "loss": 0.8699,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6864,
+      "grad_norm": 1.3746559006040033,
+      "learning_rate": 4.790559711276422e-06,
+      "loss": 0.7905,
+      "step": 1054
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 0.9627125231532317,
+      "learning_rate": 4.781257070598571e-06,
+      "loss": 0.4528,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 1.139393020426623,
+      "learning_rate": 4.771955188504371e-06,
+      "loss": 0.8002,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6912,
+      "grad_norm": 1.081193459865997,
+      "learning_rate": 4.762654097252033e-06,
+      "loss": 0.7228,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6928,
+      "grad_norm": 1.0534858365524324,
+      "learning_rate": 4.753353829097025e-06,
+      "loss": 0.6758,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 1.0009089188828206,
+      "learning_rate": 4.7440544162919645e-06,
+      "loss": 0.5022,
+      "step": 1059
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 0.9954823334743156,
+      "learning_rate": 4.734755891086498e-06,
+      "loss": 0.4838,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6976,
+      "grad_norm": 1.1670242844820142,
+      "learning_rate": 4.725458285727195e-06,
+      "loss": 0.5854,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6992,
+      "grad_norm": 1.345248711640015,
+      "learning_rate": 4.716161632457438e-06,
+      "loss": 0.7955,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7008,
+      "grad_norm": 1.0604146301112167,
+      "learning_rate": 4.7068659635173034e-06,
+      "loss": 0.6857,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 1.059910473698165,
+      "learning_rate": 4.6975713111434556e-06,
+      "loss": 0.597,
+      "step": 1064
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 1.0551783443751968,
+      "learning_rate": 4.688277707569035e-06,
+      "loss": 0.4933,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7056,
+      "grad_norm": 1.1229548669758194,
+      "learning_rate": 4.678985185023542e-06,
+      "loss": 0.5961,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7072,
+      "grad_norm": 0.9110447257531349,
+      "learning_rate": 4.669693775732733e-06,
+      "loss": 0.3572,
+      "step": 1067
+    },
+    {
+      "epoch": 1.7088,
+      "grad_norm": 1.0541361415578177,
+      "learning_rate": 4.660403511918499e-06,
+      "loss": 0.5908,
+      "step": 1068
+    },
+    {
+      "epoch": 1.7104,
+      "grad_norm": 1.3554534690356557,
+      "learning_rate": 4.65111442579876e-06,
+      "loss": 0.6871,
+      "step": 1069
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 1.1004124561307715,
+      "learning_rate": 4.641826549587352e-06,
+      "loss": 0.6668,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 1.1238666349633408,
+      "learning_rate": 4.632539915493915e-06,
+      "loss": 0.6706,
+      "step": 1071
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 1.261881791471997,
+      "learning_rate": 4.623254555723783e-06,
+      "loss": 0.6641,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7168,
+      "grad_norm": 1.007250334955659,
+      "learning_rate": 4.613970502477867e-06,
+      "loss": 0.4538,
+      "step": 1073
+    },
+    {
+      "epoch": 1.7184,
+      "grad_norm": 1.2089166770859263,
+      "learning_rate": 4.604687787952552e-06,
+      "loss": 0.7525,
+      "step": 1074
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.998155236027867,
+      "learning_rate": 4.5954064443395765e-06,
+      "loss": 0.5949,
+      "step": 1075
+    },
+    {
+      "epoch": 1.7216,
+      "grad_norm": 1.4448079202238986,
+      "learning_rate": 4.586126503825925e-06,
+      "loss": 0.7295,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 1.1810835641879265,
+      "learning_rate": 4.57684799859372e-06,
+      "loss": 0.6648,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7248,
+      "grad_norm": 1.2048129352393986,
+      "learning_rate": 4.567570960820101e-06,
+      "loss": 0.5893,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7264,
+      "grad_norm": 1.027599803617592,
+      "learning_rate": 4.558295422677124e-06,
+      "loss": 0.6012,
+      "step": 1079
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 1.1331389360529684,
+      "learning_rate": 4.54902141633164e-06,
+      "loss": 0.6808,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7296,
+      "grad_norm": 1.130502881885185,
+      "learning_rate": 4.539748973945191e-06,
+      "loss": 0.6365,
+      "step": 1081
+    },
+    {
+      "epoch": 1.7311999999999999,
+      "grad_norm": 1.1187680868884713,
+      "learning_rate": 4.53047812767389e-06,
+      "loss": 0.7399,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 1.0456949679706353,
+      "learning_rate": 4.5212089096683234e-06,
+      "loss": 0.5556,
+      "step": 1083
+    },
+    {
+      "epoch": 1.7344,
+      "grad_norm": 1.0524248612354574,
+      "learning_rate": 4.511941352073424e-06,
+      "loss": 0.6999,
+      "step": 1084
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 1.0153904119199635,
+      "learning_rate": 4.5026754870283695e-06,
+      "loss": 0.5348,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7376,
+      "grad_norm": 0.9775075407981751,
+      "learning_rate": 4.493411346666465e-06,
+      "loss": 0.4488,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7391999999999999,
+      "grad_norm": 0.972922942990288,
+      "learning_rate": 4.484148963115038e-06,
+      "loss": 0.4998,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 1.0770691912979984,
+      "learning_rate": 4.474888368495322e-06,
+      "loss": 0.6422,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 1.124485048361826,
+      "learning_rate": 4.465629594922348e-06,
+      "loss": 0.6997,
+      "step": 1089
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 1.2511248791769052,
+      "learning_rate": 4.456372674504828e-06,
+      "loss": 0.7608,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7456,
+      "grad_norm": 1.1086202746988507,
+      "learning_rate": 4.447117639345052e-06,
+      "loss": 0.6993,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7471999999999999,
+      "grad_norm": 1.04525989143107,
+      "learning_rate": 4.437864521538768e-06,
+      "loss": 0.5542,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7488000000000001,
+      "grad_norm": 0.949125942040824,
+      "learning_rate": 4.428613353175078e-06,
+      "loss": 0.541,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7504,
+      "grad_norm": 1.2102278674685583,
+      "learning_rate": 4.4193641663363214e-06,
+      "loss": 0.8433,
+      "step": 1094
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 1.1071324745488387,
+      "learning_rate": 4.410116993097968e-06,
+      "loss": 0.5821,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 1.1943502546889342,
+      "learning_rate": 4.400871865528502e-06,
+      "loss": 0.6483,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7551999999999999,
+      "grad_norm": 1.30864368936339,
+      "learning_rate": 4.391628815689314e-06,
+      "loss": 0.7407,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7568000000000001,
+      "grad_norm": 1.0196814696951229,
+      "learning_rate": 4.382387875634592e-06,
+      "loss": 0.6911,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7584,
+      "grad_norm": 1.1228657851778423,
+      "learning_rate": 4.373149077411203e-06,
+      "loss": 0.6756,
+      "step": 1099
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2581951513065248,
+      "learning_rate": 4.363912453058589e-06,
+      "loss": 0.7129,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 1.121078218340067,
+      "learning_rate": 4.354678034608654e-06,
+      "loss": 0.7299,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7631999999999999,
+      "grad_norm": 0.8367954862140153,
+      "learning_rate": 4.345445854085649e-06,
+      "loss": 0.385,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7648000000000001,
+      "grad_norm": 1.098404305635736,
+      "learning_rate": 4.336215943506066e-06,
+      "loss": 0.556,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 1.0300256778032608,
+      "learning_rate": 4.326988334878526e-06,
+      "loss": 0.4325,
+      "step": 1104
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.9839663920449834,
+      "learning_rate": 4.317763060203665e-06,
+      "loss": 0.5372,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7696,
+      "grad_norm": 0.918996899678813,
+      "learning_rate": 4.308540151474027e-06,
+      "loss": 0.4733,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 1.2009162171375496,
+      "learning_rate": 4.299319640673948e-06,
+      "loss": 0.7674,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7728000000000002,
+      "grad_norm": 0.8965525990582511,
+      "learning_rate": 4.290101559779451e-06,
+      "loss": 0.3298,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7744,
+      "grad_norm": 1.2071437169189085,
+      "learning_rate": 4.280885940758131e-06,
+      "loss": 0.6655,
+      "step": 1109
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 1.0943589484795273,
+      "learning_rate": 4.271672815569047e-06,
+      "loss": 0.6127,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7776,
+      "grad_norm": 1.100221707324751,
+      "learning_rate": 4.262462216162606e-06,
+      "loss": 0.6264,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 1.2277294029089822,
+      "learning_rate": 4.253254174480462e-06,
+      "loss": 0.772,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 1.133502627702663,
+      "learning_rate": 4.244048722455393e-06,
+      "loss": 0.6533,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7824,
+      "grad_norm": 1.132044077429827,
+      "learning_rate": 4.234845892011198e-06,
+      "loss": 0.6995,
+      "step": 1114
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 0.9962683770515747,
+      "learning_rate": 4.225645715062585e-06,
+      "loss": 0.5813,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7856,
+      "grad_norm": 1.0559637618515973,
+      "learning_rate": 4.216448223515061e-06,
+      "loss": 0.6666,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7872,
+      "grad_norm": 1.0926588684122058,
+      "learning_rate": 4.2072534492648184e-06,
+      "loss": 0.6607,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7888,
+      "grad_norm": 0.9954110873250899,
+      "learning_rate": 4.198061424198627e-06,
+      "loss": 0.6861,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 1.0381826618245467,
+      "learning_rate": 4.188872180193723e-06,
+      "loss": 0.6391,
+      "step": 1119
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 1.0918296652128439,
+      "learning_rate": 4.179685749117698e-06,
+      "loss": 0.5489,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7936,
+      "grad_norm": 1.0632920849868528,
+      "learning_rate": 4.170502162828385e-06,
+      "loss": 0.842,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7952,
+      "grad_norm": 1.1940299297769568,
+      "learning_rate": 4.161321453173759e-06,
+      "loss": 0.6134,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7968,
+      "grad_norm": 0.8993388695349045,
+      "learning_rate": 4.152143651991812e-06,
+      "loss": 0.5187,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7984,
+      "grad_norm": 1.1140215333153904,
+      "learning_rate": 4.142968791110455e-06,
+      "loss": 0.6613,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0250049674823194,
+      "learning_rate": 4.133796902347397e-06,
+      "loss": 0.4715,
+      "step": 1125
+    },
+    {
+      "epoch": 1.8016,
+      "grad_norm": 0.9374378085268168,
+      "learning_rate": 4.124628017510043e-06,
+      "loss": 0.5033,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8032,
+      "grad_norm": 1.0275824976813237,
+      "learning_rate": 4.115462168395382e-06,
+      "loss": 0.6422,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 1.250968589831042,
+      "learning_rate": 4.106299386789873e-06,
+      "loss": 0.7995,
+      "step": 1128
+    },
+    {
+      "epoch": 1.8064,
+      "grad_norm": 1.1401947903209697,
+      "learning_rate": 4.097139704469337e-06,
+      "loss": 0.7319,
+      "step": 1129
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 0.9422456049626972,
+      "learning_rate": 4.0879831531988485e-06,
+      "loss": 0.5161,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 1.1568988813017644,
+      "learning_rate": 4.078829764732621e-06,
+      "loss": 0.6733,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8112,
+      "grad_norm": 1.0840577247477663,
+      "learning_rate": 4.069679570813903e-06,
+      "loss": 0.5758,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8128,
+      "grad_norm": 0.9617887195494491,
+      "learning_rate": 4.060532603174865e-06,
+      "loss": 0.4745,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8144,
+      "grad_norm": 0.938195376837775,
+      "learning_rate": 4.051388893536484e-06,
+      "loss": 0.5375,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 1.1821054362980257,
+      "learning_rate": 4.042248473608442e-06,
+      "loss": 0.7515,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.9907229297998346,
+      "learning_rate": 4.033111375089013e-06,
+      "loss": 0.5884,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 1.1019938498630961,
+      "learning_rate": 4.023977629664951e-06,
+      "loss": 0.51,
+      "step": 1137
+    },
+    {
+      "epoch": 1.8208,
+      "grad_norm": 1.0632549365030741,
+      "learning_rate": 4.0148472690113845e-06,
+      "loss": 0.4572,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8224,
+      "grad_norm": 1.0292049071503557,
+      "learning_rate": 4.0057203247917e-06,
+      "loss": 0.7291,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 1.0827720681508033,
+      "learning_rate": 3.996596828657437e-06,
+      "loss": 0.6103,
+      "step": 1140
+    },
+    {
+      "epoch": 1.8256000000000001,
+      "grad_norm": 1.277518025094715,
+      "learning_rate": 3.987476812248181e-06,
+      "loss": 0.78,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8272,
+      "grad_norm": 0.9603646670243093,
+      "learning_rate": 3.978360307191444e-06,
+      "loss": 0.5449,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 1.1610366903286724,
+      "learning_rate": 3.969247345102567e-06,
+      "loss": 0.514,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 1.0577191345905648,
+      "learning_rate": 3.960137957584601e-06,
+      "loss": 0.6131,
+      "step": 1144
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 1.0002245927283018,
+      "learning_rate": 3.9510321762282e-06,
+      "loss": 0.59,
+      "step": 1145
+    },
+    {
+      "epoch": 1.8336000000000001,
+      "grad_norm": 0.9293742344463867,
+      "learning_rate": 3.941930032611513e-06,
+      "loss": 0.5918,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8352,
+      "grad_norm": 0.8377563940866505,
+      "learning_rate": 3.932831558300074e-06,
+      "loss": 0.4594,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8368,
+      "grad_norm": 0.9214775007622398,
+      "learning_rate": 3.923736784846693e-06,
+      "loss": 0.4994,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 1.0137162003448767,
+      "learning_rate": 3.914645743791346e-06,
+      "loss": 0.7127,
+      "step": 1149
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.8712028420975745,
+      "learning_rate": 3.90555846666106e-06,
+      "loss": 0.4343,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8416000000000001,
+      "grad_norm": 1.0595215201870019,
+      "learning_rate": 3.896474984969817e-06,
+      "loss": 0.5338,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 1.077695114776715,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.7144,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8448,
+      "grad_norm": 1.3143018246677294,
+      "learning_rate": 3.878319533894443e-06,
+      "loss": 0.7393,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8464,
+      "grad_norm": 1.0637595174147276,
+      "learning_rate": 3.869247627472021e-06,
+      "loss": 0.7207,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 1.0069524742050775,
+      "learning_rate": 3.860179642411838e-06,
+      "loss": 0.5585,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8496000000000001,
+      "grad_norm": 1.01242631348411,
+      "learning_rate": 3.851115610160967e-06,
+      "loss": 0.546,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8512,
+      "grad_norm": 1.1443497986243274,
+      "learning_rate": 3.842055562152775e-06,
+      "loss": 0.6193,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8528,
+      "grad_norm": 1.1806865043528374,
+      "learning_rate": 3.8329995298068114e-06,
+      "loss": 0.7415,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8544,
+      "grad_norm": 1.0858266668477565,
+      "learning_rate": 3.8239475445287015e-06,
+      "loss": 0.5342,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 1.2129989491446302,
+      "learning_rate": 3.814899637710031e-06,
+      "loss": 0.7803,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 1.2252042565492782,
+      "learning_rate": 3.8058558407282465e-06,
+      "loss": 0.6921,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8592,
+      "grad_norm": 1.1282891882288608,
+      "learning_rate": 3.7968161849465395e-06,
+      "loss": 0.744,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8608,
+      "grad_norm": 1.2585515340786622,
+      "learning_rate": 3.78778070171374e-06,
+      "loss": 0.7129,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8624,
+      "grad_norm": 1.0529771224424942,
+      "learning_rate": 3.7787494223642096e-06,
+      "loss": 0.5703,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 1.0938709322060731,
+      "learning_rate": 3.7697223782177304e-06,
+      "loss": 0.5712,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8656000000000001,
+      "grad_norm": 1.0605176733000814,
+      "learning_rate": 3.760699600579399e-06,
+      "loss": 0.6075,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 1.0363584586659609,
+      "learning_rate": 3.7516811207395116e-06,
+      "loss": 0.6045,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 1.142411464820728,
+      "learning_rate": 3.742666969973463e-06,
+      "loss": 0.7302,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8704,
+      "grad_norm": 1.0954726085406266,
+      "learning_rate": 3.733657179541635e-06,
+      "loss": 0.6496,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 1.0629301918194092,
+      "learning_rate": 3.724651780689286e-06,
+      "loss": 0.6655,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8736000000000002,
+      "grad_norm": 1.0675645700003733,
+      "learning_rate": 3.715650804646449e-06,
+      "loss": 0.6425,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8752,
+      "grad_norm": 0.9670180398783638,
+      "learning_rate": 3.7066542826278153e-06,
+      "loss": 0.5094,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 1.0924204822663228,
+      "learning_rate": 3.6976622458326308e-06,
+      "loss": 0.7368,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8784,
+      "grad_norm": 1.3398593468593558,
+      "learning_rate": 3.6886747254445877e-06,
+      "loss": 0.8557,
+      "step": 1174
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1046487320646705,
+      "learning_rate": 3.6796917526317153e-06,
+      "loss": 0.6763,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 1.0936990912044038,
+      "learning_rate": 3.6707133585462713e-06,
+      "loss": 0.5341,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8832,
+      "grad_norm": 1.232437486300085,
+      "learning_rate": 3.6617395743246375e-06,
+      "loss": 0.7537,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8848,
+      "grad_norm": 1.0203715444637904,
+      "learning_rate": 3.652770431087206e-06,
+      "loss": 0.6408,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 1.0082403503624886,
+      "learning_rate": 3.6438059599382765e-06,
+      "loss": 0.5113,
+      "step": 1179
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 0.9959890812617046,
+      "learning_rate": 3.634846191965944e-06,
+      "loss": 0.699,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8896,
+      "grad_norm": 1.1058286048074997,
+      "learning_rate": 3.625891158241994e-06,
+      "loss": 0.6435,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8912,
+      "grad_norm": 0.969966661507889,
+      "learning_rate": 3.6169408898217973e-06,
+      "loss": 0.6663,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8928,
+      "grad_norm": 0.8505764415202092,
+      "learning_rate": 3.6079954177441945e-06,
+      "loss": 0.309,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 1.0758446208904986,
+      "learning_rate": 3.599054773031394e-06,
+      "loss": 0.5717,
+      "step": 1184
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 1.0165347761828614,
+      "learning_rate": 3.5901189866888654e-06,
+      "loss": 0.4677,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8976,
+      "grad_norm": 0.9286025334800835,
+      "learning_rate": 3.581188089705226e-06,
+      "loss": 0.4619,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8992,
+      "grad_norm": 0.9374588042066216,
+      "learning_rate": 3.572262113052142e-06,
+      "loss": 0.3228,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9008,
+      "grad_norm": 1.025736743071331,
+      "learning_rate": 3.563341087684213e-06,
+      "loss": 0.6244,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9024,
+      "grad_norm": 1.1222808943789733,
+      "learning_rate": 3.554425044538868e-06,
+      "loss": 0.6511,
+      "step": 1189
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 1.1996286960494713,
+      "learning_rate": 3.5455140145362587e-06,
+      "loss": 0.6727,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 1.149147491424003,
+      "learning_rate": 3.5366080285791516e-06,
+      "loss": 0.5308,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.8585098677052776,
+      "learning_rate": 3.527707117552822e-06,
+      "loss": 0.4403,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9088,
+      "grad_norm": 0.9803858690486458,
+      "learning_rate": 3.5188113123249435e-06,
+      "loss": 0.475,
+      "step": 1193
+    },
+    {
+      "epoch": 1.9104,
+      "grad_norm": 1.1273091414872751,
+      "learning_rate": 3.5099206437454852e-06,
+      "loss": 0.8288,
+      "step": 1194
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 0.9677683578490914,
+      "learning_rate": 3.5010351426466006e-06,
+      "loss": 0.5636,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9136,
+      "grad_norm": 1.0976673251534452,
+      "learning_rate": 3.4921548398425246e-06,
+      "loss": 0.6977,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 1.289293306657935,
+      "learning_rate": 3.4832797661294633e-06,
+      "loss": 0.8418,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9167999999999998,
+      "grad_norm": 1.1176129391055853,
+      "learning_rate": 3.4744099522854914e-06,
+      "loss": 0.5438,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9184,
+      "grad_norm": 1.0622579906717327,
+      "learning_rate": 3.4655454290704393e-06,
+      "loss": 0.6945,
+      "step": 1199
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.143160951662353,
+      "learning_rate": 3.4566862272257923e-06,
+      "loss": 0.6156,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9216,
+      "grad_norm": 1.1044820527219912,
+      "learning_rate": 3.44783237747458e-06,
+      "loss": 0.6793,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9232,
+      "grad_norm": 0.913587198928814,
+      "learning_rate": 3.438983910521273e-06,
+      "loss": 0.4574,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 1.0262417303481601,
+      "learning_rate": 3.430140857051675e-06,
+      "loss": 0.524,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9264000000000001,
+      "grad_norm": 1.0448064318747774,
+      "learning_rate": 3.421303247732813e-06,
+      "loss": 0.622,
+      "step": 1204
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.9968936147775568,
+      "learning_rate": 3.4124711132128374e-06,
+      "loss": 0.6706,
+      "step": 1205
+    },
+    {
+      "epoch": 1.9296,
+      "grad_norm": 1.0266056364903204,
+      "learning_rate": 3.4036444841209113e-06,
+      "loss": 0.5003,
+      "step": 1206
+    },
+    {
+      "epoch": 1.9312,
+      "grad_norm": 1.0229524282335964,
+      "learning_rate": 3.3948233910671036e-06,
+      "loss": 0.5934,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 1.0464722207776467,
+      "learning_rate": 3.3860078646422894e-06,
+      "loss": 0.6611,
+      "step": 1208
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 1.272725949443715,
+      "learning_rate": 3.3771979354180343e-06,
+      "loss": 0.8157,
+      "step": 1209
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 1.1164579964923933,
+      "learning_rate": 3.3683936339464957e-06,
+      "loss": 0.6552,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9376,
+      "grad_norm": 1.0948215198483193,
+      "learning_rate": 3.359594990760313e-06,
+      "loss": 0.7712,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9392,
+      "grad_norm": 0.9259793414482724,
+      "learning_rate": 3.3508020363725043e-06,
+      "loss": 0.4601,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9407999999999999,
+      "grad_norm": 1.0148943617891273,
+      "learning_rate": 3.34201480127636e-06,
+      "loss": 0.577,
+      "step": 1213
+    },
+    {
+      "epoch": 1.9424000000000001,
+      "grad_norm": 1.0599539657119323,
+      "learning_rate": 3.333233315945333e-06,
+      "loss": 0.6929,
+      "step": 1214
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 1.111548261387851,
+      "learning_rate": 3.324457610832942e-06,
+      "loss": 0.6392,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.9693411447002284,
+      "learning_rate": 3.315687716372655e-06,
+      "loss": 0.602,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9472,
+      "grad_norm": 1.0491414415783453,
+      "learning_rate": 3.306923662977789e-06,
+      "loss": 0.6368,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9487999999999999,
+      "grad_norm": 1.2755172089583906,
+      "learning_rate": 3.2981654810414128e-06,
+      "loss": 0.7871,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9504000000000001,
+      "grad_norm": 1.1889427686928447,
+      "learning_rate": 3.2894132009362245e-06,
+      "loss": 0.7055,
+      "step": 1219
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 1.0512550390583228,
+      "learning_rate": 3.280666853014457e-06,
+      "loss": 0.7371,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 1.1053457261087656,
+      "learning_rate": 3.271926467607774e-06,
+      "loss": 0.9201,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9552,
+      "grad_norm": 0.764537316486806,
+      "learning_rate": 3.2631920750271594e-06,
+      "loss": 0.299,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9567999999999999,
+      "grad_norm": 1.1013151028500678,
+      "learning_rate": 3.2544637055628135e-06,
+      "loss": 0.7675,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 1.1106418445394894,
+      "learning_rate": 3.2457413894840516e-06,
+      "loss": 0.7402,
+      "step": 1224
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9517371660120999,
+      "learning_rate": 3.2370251570391933e-06,
+      "loss": 0.5603,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9616,
+      "grad_norm": 0.9750655566652322,
+      "learning_rate": 3.2283150384554642e-06,
+      "loss": 0.5366,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 1.0126524621674864,
+      "learning_rate": 3.219611063938883e-06,
+      "loss": 0.6454,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9647999999999999,
+      "grad_norm": 1.0848702741790708,
+      "learning_rate": 3.210913263674166e-06,
+      "loss": 0.8955,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9664000000000001,
+      "grad_norm": 1.2630036580277868,
+      "learning_rate": 3.2022216678246145e-06,
+      "loss": 0.8736,
+      "step": 1229
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 0.9905793845038109,
+      "learning_rate": 3.193536306532013e-06,
+      "loss": 0.6744,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9696,
+      "grad_norm": 1.1625027336896117,
+      "learning_rate": 3.184857209916528e-06,
+      "loss": 0.7322,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.9371978339570802,
+      "learning_rate": 3.1761844080765993e-06,
+      "loss": 0.5515,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 1.1838034658558134,
+      "learning_rate": 3.1675179310888344e-06,
+      "loss": 0.7449,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9744000000000002,
+      "grad_norm": 1.0231905064327331,
+      "learning_rate": 3.15885780900791e-06,
+      "loss": 0.5697,
+      "step": 1234
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 1.1284739741057463,
+      "learning_rate": 3.150204071866464e-06,
+      "loss": 0.7329,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9776,
+      "grad_norm": 0.9501799272617544,
+      "learning_rate": 3.141556749674988e-06,
+      "loss": 0.5635,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9792,
+      "grad_norm": 0.9645096482859683,
+      "learning_rate": 3.132915872421734e-06,
+      "loss": 0.5907,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9808,
+      "grad_norm": 0.9657249403382168,
+      "learning_rate": 3.1242814700725977e-06,
+      "loss": 0.507,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 1.1876940629033004,
+      "learning_rate": 3.1156535725710224e-06,
+      "loss": 0.7031,
+      "step": 1239
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.8677647270380617,
+      "learning_rate": 3.1070322098378925e-06,
+      "loss": 0.5107,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9856,
+      "grad_norm": 1.0104654552308285,
+      "learning_rate": 3.0984174117714306e-06,
+      "loss": 0.6148,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9872,
+      "grad_norm": 1.0403754482592862,
+      "learning_rate": 3.0898092082470943e-06,
+      "loss": 0.6376,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9888,
+      "grad_norm": 0.9512295950257441,
+      "learning_rate": 3.081207629117472e-06,
+      "loss": 0.5417,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9904,
+      "grad_norm": 1.0660226463461617,
+      "learning_rate": 3.0726127042121766e-06,
+      "loss": 0.6791,
+      "step": 1244
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 1.0911812683238395,
+      "learning_rate": 3.064024463337747e-06,
+      "loss": 0.6689,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9936,
+      "grad_norm": 1.096816618636096,
+      "learning_rate": 3.0554429362775417e-06,
+      "loss": 0.5493,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9952,
+      "grad_norm": 1.0170905967027304,
+      "learning_rate": 3.046868152791638e-06,
+      "loss": 0.5411,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 1.1719500927471784,
+      "learning_rate": 3.038300142616723e-06,
+      "loss": 0.7876,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9984,
+      "grad_norm": 1.0519813140461414,
+      "learning_rate": 3.0297389354659984e-06,
+      "loss": 0.5307,
+      "step": 1249
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9812973860553407,
+      "learning_rate": 3.021184561029071e-06,
+      "loss": 0.5712,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0016,
+      "grad_norm": 1.298399167242736,
+      "learning_rate": 3.0126370489718537e-06,
+      "loss": 0.8351,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0032,
+      "grad_norm": 1.0115570091366153,
+      "learning_rate": 3.0040964289364618e-06,
+      "loss": 0.5913,
+      "step": 1252
+    },
+    {
+      "epoch": 2.0048,
+      "grad_norm": 0.8641326830827756,
+      "learning_rate": 2.9955627305411074e-06,
+      "loss": 0.3653,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.730105823623971,
+      "learning_rate": 2.9870359833799994e-06,
+      "loss": 0.325,
+      "step": 1254
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 1.0682839585337933,
+      "learning_rate": 2.978516217023243e-06,
+      "loss": 0.6459,
+      "step": 1255
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 1.0124894026880333,
+      "learning_rate": 2.97000346101673e-06,
+      "loss": 0.4041,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0112,
+      "grad_norm": 0.9836866148081452,
+      "learning_rate": 2.9614977448820444e-06,
+      "loss": 0.6355,
+      "step": 1257
+    },
+    {
+      "epoch": 2.0128,
+      "grad_norm": 0.8881484875809239,
+      "learning_rate": 2.952999098116356e-06,
+      "loss": 0.4098,
+      "step": 1258
+    },
+    {
+      "epoch": 2.0144,
+      "grad_norm": 0.9034890708378922,
+      "learning_rate": 2.944507550192318e-06,
+      "loss": 0.4942,
+      "step": 1259
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 1.080783083458099,
+      "learning_rate": 2.9360231305579645e-06,
+      "loss": 0.6983,
+      "step": 1260
+    },
+    {
+      "epoch": 2.0176,
+      "grad_norm": 1.0559845630378675,
+      "learning_rate": 2.9275458686366108e-06,
+      "loss": 0.6482,
+      "step": 1261
+    },
+    {
+      "epoch": 2.0192,
+      "grad_norm": 0.9043201051653957,
+      "learning_rate": 2.9190757938267477e-06,
+      "loss": 0.4773,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0208,
+      "grad_norm": 0.9566088998528108,
+      "learning_rate": 2.9106129355019464e-06,
+      "loss": 0.4833,
+      "step": 1263
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.7878589615240992,
+      "learning_rate": 2.9021573230107436e-06,
+      "loss": 0.3616,
+      "step": 1264
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 0.992823396551406,
+      "learning_rate": 2.8937089856765564e-06,
+      "loss": 0.4359,
+      "step": 1265
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 1.0795739393365553,
+      "learning_rate": 2.885267952797569e-06,
+      "loss": 0.686,
+      "step": 1266
+    },
+    {
+      "epoch": 2.0272,
+      "grad_norm": 1.0291256365245904,
+      "learning_rate": 2.876834253646631e-06,
+      "loss": 0.5252,
+      "step": 1267
+    },
+    {
+      "epoch": 2.0288,
+      "grad_norm": 0.7232351482379552,
+      "learning_rate": 2.8684079174711665e-06,
+      "loss": 0.3315,
+      "step": 1268
+    },
+    {
+      "epoch": 2.0304,
+      "grad_norm": 1.0232833436538669,
+      "learning_rate": 2.8599889734930548e-06,
+      "loss": 0.5851,
+      "step": 1269
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 1.073593131217862,
+      "learning_rate": 2.8515774509085535e-06,
+      "loss": 0.6536,
+      "step": 1270
+    },
+    {
+      "epoch": 2.0336,
+      "grad_norm": 0.980618720389753,
+      "learning_rate": 2.8431733788881703e-06,
+      "loss": 0.4863,
+      "step": 1271
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 1.1942363095739834,
+      "learning_rate": 2.8347767865765828e-06,
+      "loss": 0.8438,
+      "step": 1272
+    },
+    {
+      "epoch": 2.0368,
+      "grad_norm": 1.0461300334530252,
+      "learning_rate": 2.826387703092528e-06,
+      "loss": 0.635,
+      "step": 1273
+    },
+    {
+      "epoch": 2.0384,
+      "grad_norm": 0.9422807925717059,
+      "learning_rate": 2.8180061575286995e-06,
+      "loss": 0.5464,
+      "step": 1274
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.9678534430805344,
+      "learning_rate": 2.8096321789516557e-06,
+      "loss": 0.5721,
+      "step": 1275
+    },
+    {
+      "epoch": 2.0416,
+      "grad_norm": 1.0140477750718446,
+      "learning_rate": 2.8012657964017056e-06,
+      "loss": 0.573,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0432,
+      "grad_norm": 0.9698324585025372,
+      "learning_rate": 2.792907038892823e-06,
+      "loss": 0.4951,
+      "step": 1277
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.9245838432325915,
+      "learning_rate": 2.784555935412538e-06,
+      "loss": 0.5445,
+      "step": 1278
+    },
+    {
+      "epoch": 2.0464,
+      "grad_norm": 0.8639851249600945,
+      "learning_rate": 2.77621251492183e-06,
+      "loss": 0.417,
+      "step": 1279
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.8447457003746222,
+      "learning_rate": 2.7678768063550454e-06,
+      "loss": 0.3163,
+      "step": 1280
+    },
+    {
+      "epoch": 2.0496,
+      "grad_norm": 0.9076286544281497,
+      "learning_rate": 2.759548838619774e-06,
+      "loss": 0.5033,
+      "step": 1281
+    },
+    {
+      "epoch": 2.0512,
+      "grad_norm": 1.2366268110917877,
+      "learning_rate": 2.7512286405967726e-06,
+      "loss": 0.6157,
+      "step": 1282
+    },
+    {
+      "epoch": 2.0528,
+      "grad_norm": 0.9600144959233128,
+      "learning_rate": 2.742916241139843e-06,
+      "loss": 0.4675,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.8257950340804546,
+      "learning_rate": 2.7346116690757496e-06,
+      "loss": 0.393,
+      "step": 1284
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 0.858074362985321,
+      "learning_rate": 2.726314953204111e-06,
+      "loss": 0.3556,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0576,
+      "grad_norm": 1.1331858093637908,
+      "learning_rate": 2.718026122297297e-06,
+      "loss": 0.6141,
+      "step": 1286
+    },
+    {
+      "epoch": 2.0592,
+      "grad_norm": 0.9248025325714058,
+      "learning_rate": 2.7097452051003375e-06,
+      "loss": 0.5012,
+      "step": 1287
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 1.1921062119538375,
+      "learning_rate": 2.701472230330813e-06,
+      "loss": 0.5427,
+      "step": 1288
+    },
+    {
+      "epoch": 2.0624,
+      "grad_norm": 1.1033206461763625,
+      "learning_rate": 2.693207226678767e-06,
+      "loss": 0.7022,
+      "step": 1289
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.8214308375608488,
+      "learning_rate": 2.684950222806596e-06,
+      "loss": 0.4938,
+      "step": 1290
+    },
+    {
+      "epoch": 2.0656,
+      "grad_norm": 0.9998145252573861,
+      "learning_rate": 2.676701247348951e-06,
+      "loss": 0.5156,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0672,
+      "grad_norm": 1.0604065887698322,
+      "learning_rate": 2.6684603289126492e-06,
+      "loss": 0.8088,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0688,
+      "grad_norm": 0.7965545543580673,
+      "learning_rate": 2.660227496076557e-06,
+      "loss": 0.3896,
+      "step": 1293
+    },
+    {
+      "epoch": 2.0704,
+      "grad_norm": 1.0123116241883714,
+      "learning_rate": 2.6520027773915075e-06,
+      "loss": 0.6165,
+      "step": 1294
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 0.9406371451075433,
+      "learning_rate": 2.643786201380194e-06,
+      "loss": 0.5054,
+      "step": 1295
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 1.0095624346960441,
+      "learning_rate": 2.6355777965370665e-06,
+      "loss": 0.6304,
+      "step": 1296
+    },
+    {
+      "epoch": 2.0752,
+      "grad_norm": 1.0610164240831368,
+      "learning_rate": 2.627377591328245e-06,
+      "loss": 0.6221,
+      "step": 1297
+    },
+    {
+      "epoch": 2.0768,
+      "grad_norm": 1.0192376687862645,
+      "learning_rate": 2.6191856141914074e-06,
+      "loss": 0.5953,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0784,
+      "grad_norm": 0.8391614462784376,
+      "learning_rate": 2.6110018935357005e-06,
+      "loss": 0.5315,
+      "step": 1299
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.9016029727380366,
+      "learning_rate": 2.6028264577416418e-06,
+      "loss": 0.5358,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0816,
+      "grad_norm": 0.8749253164294157,
+      "learning_rate": 2.594659335161008e-06,
+      "loss": 0.4995,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.8434346505653489,
+      "learning_rate": 2.586500554116757e-06,
+      "loss": 0.2961,
+      "step": 1302
+    },
+    {
+      "epoch": 2.0848,
+      "grad_norm": 0.7870803150945824,
+      "learning_rate": 2.578350142902909e-06,
+      "loss": 0.3278,
+      "step": 1303
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 1.1245913686835163,
+      "learning_rate": 2.570208129784466e-06,
+      "loss": 0.6175,
+      "step": 1304
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 0.9544970911326556,
+      "learning_rate": 2.562074542997305e-06,
+      "loss": 0.4934,
+      "step": 1305
+    },
+    {
+      "epoch": 2.0896,
+      "grad_norm": 0.8512076363328573,
+      "learning_rate": 2.5539494107480746e-06,
+      "loss": 0.4864,
+      "step": 1306
+    },
+    {
+      "epoch": 2.0912,
+      "grad_norm": 0.8888992444824483,
+      "learning_rate": 2.545832761214112e-06,
+      "loss": 0.309,
+      "step": 1307
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.9266980416401082,
+      "learning_rate": 2.5377246225433306e-06,
+      "loss": 0.3737,
+      "step": 1308
+    },
+    {
+      "epoch": 2.0944,
+      "grad_norm": 0.9592955489559961,
+      "learning_rate": 2.529625022854133e-06,
+      "loss": 0.5452,
+      "step": 1309
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 0.9548284333007332,
+      "learning_rate": 2.5215339902353097e-06,
+      "loss": 0.488,
+      "step": 1310
+    },
+    {
+      "epoch": 2.0976,
+      "grad_norm": 0.9815663671906946,
+      "learning_rate": 2.513451552745936e-06,
+      "loss": 0.5634,
+      "step": 1311
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.9452581004519367,
+      "learning_rate": 2.505377738415286e-06,
+      "loss": 0.5304,
+      "step": 1312
+    },
+    {
+      "epoch": 2.1008,
+      "grad_norm": 1.0235311267996225,
+      "learning_rate": 2.4973125752427243e-06,
+      "loss": 0.7051,
+      "step": 1313
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.8689338688886353,
+      "learning_rate": 2.4892560911976167e-06,
+      "loss": 0.3737,
+      "step": 1314
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 0.7522714248452472,
+      "learning_rate": 2.481208314219233e-06,
+      "loss": 0.3891,
+      "step": 1315
+    },
+    {
+      "epoch": 2.1056,
+      "grad_norm": 0.9506359372253496,
+      "learning_rate": 2.4731692722166408e-06,
+      "loss": 0.4796,
+      "step": 1316
+    },
+    {
+      "epoch": 2.1072,
+      "grad_norm": 0.9711766706896126,
+      "learning_rate": 2.4651389930686226e-06,
+      "loss": 0.6058,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1088,
+      "grad_norm": 0.9037619230148777,
+      "learning_rate": 2.457117504623565e-06,
+      "loss": 0.4974,
+      "step": 1318
+    },
+    {
+      "epoch": 2.1104,
+      "grad_norm": 1.0168956995397698,
+      "learning_rate": 2.4491048346993756e-06,
+      "loss": 0.6248,
+      "step": 1319
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6190604609192364,
+      "learning_rate": 2.4411010110833783e-06,
+      "loss": 0.245,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1136,
+      "grad_norm": 0.8802450641781031,
+      "learning_rate": 2.433106061532214e-06,
+      "loss": 0.3597,
+      "step": 1321
+    },
+    {
+      "epoch": 2.1152,
+      "grad_norm": 0.8716063847557931,
+      "learning_rate": 2.4251200137717545e-06,
+      "loss": 0.3304,
+      "step": 1322
+    },
+    {
+      "epoch": 2.1168,
+      "grad_norm": 0.9272577022059759,
+      "learning_rate": 2.4171428954969995e-06,
+      "loss": 0.5488,
+      "step": 1323
+    },
+    {
+      "epoch": 2.1184,
+      "grad_norm": 1.3417573867035006,
+      "learning_rate": 2.4091747343719828e-06,
+      "loss": 0.6259,
+      "step": 1324
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8949851079943691,
+      "learning_rate": 2.401215558029671e-06,
+      "loss": 0.4687,
+      "step": 1325
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.9844009107288849,
+      "learning_rate": 2.3932653940718784e-06,
+      "loss": 0.4502,
+      "step": 1326
+    },
+    {
+      "epoch": 2.1232,
+      "grad_norm": 0.9943279358277654,
+      "learning_rate": 2.3853242700691594e-06,
+      "loss": 0.4865,
+      "step": 1327
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.8693025852066528,
+      "learning_rate": 2.3773922135607217e-06,
+      "loss": 0.3432,
+      "step": 1328
+    },
+    {
+      "epoch": 2.1264,
+      "grad_norm": 1.0288277841368596,
+      "learning_rate": 2.3694692520543293e-06,
+      "loss": 0.5083,
+      "step": 1329
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 0.975241454721127,
+      "learning_rate": 2.3615554130262003e-06,
+      "loss": 0.5066,
+      "step": 1330
+    },
+    {
+      "epoch": 2.1296,
+      "grad_norm": 1.0101395736110435,
+      "learning_rate": 2.3536507239209223e-06,
+      "loss": 0.4061,
+      "step": 1331
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.8846159887382024,
+      "learning_rate": 2.3457552121513455e-06,
+      "loss": 0.4469,
+      "step": 1332
+    },
+    {
+      "epoch": 2.1328,
+      "grad_norm": 1.1805939060907864,
+      "learning_rate": 2.337868905098499e-06,
+      "loss": 0.753,
+      "step": 1333
+    },
+    {
+      "epoch": 2.1344,
+      "grad_norm": 0.950570922859334,
+      "learning_rate": 2.329991830111492e-06,
+      "loss": 0.4998,
+      "step": 1334
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 1.1325339806766936,
+      "learning_rate": 2.32212401450741e-06,
+      "loss": 0.6582,
+      "step": 1335
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 1.1345026541207857,
+      "learning_rate": 2.3142654855712353e-06,
+      "loss": 0.6846,
+      "step": 1336
+    },
+    {
+      "epoch": 2.1391999999999998,
+      "grad_norm": 0.9142443518199547,
+      "learning_rate": 2.3064162705557387e-06,
+      "loss": 0.4974,
+      "step": 1337
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 1.2171039149029688,
+      "learning_rate": 2.2985763966813963e-06,
+      "loss": 0.6297,
+      "step": 1338
+    },
+    {
+      "epoch": 2.1424,
+      "grad_norm": 1.103528686330069,
+      "learning_rate": 2.2907458911362885e-06,
+      "loss": 0.5007,
+      "step": 1339
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 0.98275498808103,
+      "learning_rate": 2.2829247810760023e-06,
+      "loss": 0.604,
+      "step": 1340
+    },
+    {
+      "epoch": 2.1456,
+      "grad_norm": 1.14920374586163,
+      "learning_rate": 2.275113093623551e-06,
+      "loss": 0.6044,
+      "step": 1341
+    },
+    {
+      "epoch": 2.1471999999999998,
+      "grad_norm": 0.8895248139503149,
+      "learning_rate": 2.2673108558692603e-06,
+      "loss": 0.5106,
+      "step": 1342
+    },
+    {
+      "epoch": 2.1488,
+      "grad_norm": 1.0383979371709695,
+      "learning_rate": 2.259518094870693e-06,
+      "loss": 0.6747,
+      "step": 1343
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.9837990098662202,
+      "learning_rate": 2.251734837652547e-06,
+      "loss": 0.5293,
+      "step": 1344
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 0.9162577478720801,
+      "learning_rate": 2.243961111206555e-06,
+      "loss": 0.5038,
+      "step": 1345
+    },
+    {
+      "epoch": 2.1536,
+      "grad_norm": 0.9406323194771262,
+      "learning_rate": 2.236196942491407e-06,
+      "loss": 0.5839,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1552,
+      "grad_norm": 0.9168585279393038,
+      "learning_rate": 2.228442358432638e-06,
+      "loss": 0.5146,
+      "step": 1347
+    },
+    {
+      "epoch": 2.1568,
+      "grad_norm": 0.8734762641042678,
+      "learning_rate": 2.2206973859225518e-06,
+      "loss": 0.4854,
+      "step": 1348
+    },
+    {
+      "epoch": 2.1584,
+      "grad_norm": 0.9192358210274748,
+      "learning_rate": 2.2129620518201184e-06,
+      "loss": 0.5738,
+      "step": 1349
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.9260606627189742,
+      "learning_rate": 2.2052363829508776e-06,
+      "loss": 0.5088,
+      "step": 1350
+    },
+    {
+      "epoch": 2.1616,
+      "grad_norm": 0.9875829468141519,
+      "learning_rate": 2.1975204061068594e-06,
+      "loss": 0.6714,
+      "step": 1351
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 1.009298443675721,
+      "learning_rate": 2.189814148046473e-06,
+      "loss": 0.6083,
+      "step": 1352
+    },
+    {
+      "epoch": 2.1648,
+      "grad_norm": 0.8948995116074715,
+      "learning_rate": 2.182117635494431e-06,
+      "loss": 0.4535,
+      "step": 1353
+    },
+    {
+      "epoch": 2.1664,
+      "grad_norm": 0.8766303225848553,
+      "learning_rate": 2.1744308951416483e-06,
+      "loss": 0.5383,
+      "step": 1354
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 0.8576291091780098,
+      "learning_rate": 2.1667539536451455e-06,
+      "loss": 0.4832,
+      "step": 1355
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.8571069713742667,
+      "learning_rate": 2.1590868376279693e-06,
+      "loss": 0.4328,
+      "step": 1356
+    },
+    {
+      "epoch": 2.1712,
+      "grad_norm": 1.0630673088432463,
+      "learning_rate": 2.151429573679084e-06,
+      "loss": 0.6379,
+      "step": 1357
+    },
+    {
+      "epoch": 2.1728,
+      "grad_norm": 1.0078714612201245,
+      "learning_rate": 2.1437821883532956e-06,
+      "loss": 0.3368,
+      "step": 1358
+    },
+    {
+      "epoch": 2.1744,
+      "grad_norm": 1.07940454825296,
+      "learning_rate": 2.136144708171145e-06,
+      "loss": 0.5254,
+      "step": 1359
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 1.0133271267795005,
+      "learning_rate": 2.128517159618827e-06,
+      "loss": 0.7062,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1776,
+      "grad_norm": 0.8300476112126882,
+      "learning_rate": 2.1208995691480947e-06,
+      "loss": 0.3922,
+      "step": 1361
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.7446518783717817,
+      "learning_rate": 2.1132919631761637e-06,
+      "loss": 0.3887,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1808,
+      "grad_norm": 0.9550817643897436,
+      "learning_rate": 2.1056943680856286e-06,
+      "loss": 0.5336,
+      "step": 1363
+    },
+    {
+      "epoch": 2.1824,
+      "grad_norm": 0.9793688986781,
+      "learning_rate": 2.098106810224362e-06,
+      "loss": 0.3602,
+      "step": 1364
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 1.1524472163611443,
+      "learning_rate": 2.0905293159054315e-06,
+      "loss": 0.5893,
+      "step": 1365
+    },
+    {
+      "epoch": 2.1856,
+      "grad_norm": 0.8804131590200853,
+      "learning_rate": 2.0829619114070068e-06,
+      "loss": 0.5328,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1872,
+      "grad_norm": 1.0627916866142488,
+      "learning_rate": 2.075404622972261e-06,
+      "loss": 0.5919,
+      "step": 1367
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 1.004042558792755,
+      "learning_rate": 2.0678574768092926e-06,
+      "loss": 0.5994,
+      "step": 1368
+    },
+    {
+      "epoch": 2.1904,
+      "grad_norm": 1.1220282135212183,
+      "learning_rate": 2.0603204990910195e-06,
+      "loss": 0.6346,
+      "step": 1369
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 1.0007162097947635,
+      "learning_rate": 2.0527937159551044e-06,
+      "loss": 0.5406,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1936,
+      "grad_norm": 1.047637556299616,
+      "learning_rate": 2.0452771535038518e-06,
+      "loss": 0.6951,
+      "step": 1371
+    },
+    {
+      "epoch": 2.1952,
+      "grad_norm": 0.8857741729314739,
+      "learning_rate": 2.03777083780412e-06,
+      "loss": 0.4623,
+      "step": 1372
+    },
+    {
+      "epoch": 2.1968,
+      "grad_norm": 1.0427427424808937,
+      "learning_rate": 2.030274794887237e-06,
+      "loss": 0.6872,
+      "step": 1373
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.7469896916379233,
+      "learning_rate": 2.0227890507488993e-06,
+      "loss": 0.3957,
+      "step": 1374
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9920019294995017,
+      "learning_rate": 2.0153136313490945e-06,
+      "loss": 0.5796,
+      "step": 1375
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.8442658679814324,
+      "learning_rate": 2.0078485626120015e-06,
+      "loss": 0.4201,
+      "step": 1376
+    },
+    {
+      "epoch": 2.2032,
+      "grad_norm": 1.0650446766943666,
+      "learning_rate": 2.000393870425904e-06,
+      "loss": 0.6618,
+      "step": 1377
+    },
+    {
+      "epoch": 2.2048,
+      "grad_norm": 1.2360614868755886,
+      "learning_rate": 1.9929495806431024e-06,
+      "loss": 0.69,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2064,
+      "grad_norm": 0.9834614831106662,
+      "learning_rate": 1.985515719079819e-06,
+      "loss": 0.5347,
+      "step": 1379
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.9563403918912294,
+      "learning_rate": 1.978092311516116e-06,
+      "loss": 0.573,
+      "step": 1380
+    },
+    {
+      "epoch": 2.2096,
+      "grad_norm": 0.8929863110801631,
+      "learning_rate": 1.9706793836957964e-06,
+      "loss": 0.4499,
+      "step": 1381
+    },
+    {
+      "epoch": 2.2112,
+      "grad_norm": 0.9649251196073643,
+      "learning_rate": 1.963276961326326e-06,
+      "loss": 0.5692,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2128,
+      "grad_norm": 0.9964642188265687,
+      "learning_rate": 1.955885070078737e-06,
+      "loss": 0.5104,
+      "step": 1383
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.8444129388269784,
+      "learning_rate": 1.948503735587537e-06,
+      "loss": 0.4842,
+      "step": 1384
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 1.0246292975373192,
+      "learning_rate": 1.9411329834506286e-06,
+      "loss": 0.6456,
+      "step": 1385
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.9447698744612065,
+      "learning_rate": 1.9337728392292104e-06,
+      "loss": 0.5896,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2192,
+      "grad_norm": 0.9486087997092717,
+      "learning_rate": 1.926423328447698e-06,
+      "loss": 0.4815,
+      "step": 1387
+    },
+    {
+      "epoch": 2.2208,
+      "grad_norm": 0.9007389003222603,
+      "learning_rate": 1.919084476593631e-06,
+      "loss": 0.4841,
+      "step": 1388
+    },
+    {
+      "epoch": 2.2224,
+      "grad_norm": 1.1009667113474817,
+      "learning_rate": 1.9117563091175795e-06,
+      "loss": 0.618,
+      "step": 1389
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 0.8885485098677216,
+      "learning_rate": 1.904438851433068e-06,
+      "loss": 0.4859,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2256,
+      "grad_norm": 0.9007315166212934,
+      "learning_rate": 1.897132128916474e-06,
+      "loss": 0.5454,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.9547852523963741,
+      "learning_rate": 1.8898361669069497e-06,
+      "loss": 0.5328,
+      "step": 1392
+    },
+    {
+      "epoch": 2.2288,
+      "grad_norm": 1.0780777940516166,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.5198,
+      "step": 1393
+    },
+    {
+      "epoch": 2.2304,
+      "grad_norm": 1.104429607731545,
+      "learning_rate": 1.87527662557905e-06,
+      "loss": 0.7858,
+      "step": 1394
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 0.9146270303412359,
+      "learning_rate": 1.8680130967520433e-06,
+      "loss": 0.5381,
+      "step": 1395
+    },
+    {
+      "epoch": 2.2336,
+      "grad_norm": 0.9703383255654834,
+      "learning_rate": 1.8607604294146685e-06,
+      "loss": 0.4282,
+      "step": 1396
+    },
+    {
+      "epoch": 2.2352,
+      "grad_norm": 0.8865861347723203,
+      "learning_rate": 1.8535186487186213e-06,
+      "loss": 0.4508,
+      "step": 1397
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.9067456819974662,
+      "learning_rate": 1.8462877797778367e-06,
+      "loss": 0.589,
+      "step": 1398
+    },
+    {
+      "epoch": 2.2384,
+      "grad_norm": 1.012852254845857,
+      "learning_rate": 1.8390678476684143e-06,
+      "loss": 0.6066,
+      "step": 1399
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.8746088320986213,
+      "learning_rate": 1.831858877428524e-06,
+      "loss": 0.5811,
+      "step": 1400
+    },
+    {
+      "epoch": 2.2416,
+      "grad_norm": 0.7862613026231213,
+      "learning_rate": 1.8246608940583166e-06,
+      "loss": 0.4263,
+      "step": 1401
+    },
+    {
+      "epoch": 2.2432,
+      "grad_norm": 0.8646940392936836,
+      "learning_rate": 1.8174739225198485e-06,
+      "loss": 0.4494,
+      "step": 1402
+    },
+    {
+      "epoch": 2.2448,
+      "grad_norm": 1.024426874039442,
+      "learning_rate": 1.8102979877369808e-06,
+      "loss": 0.5053,
+      "step": 1403
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.9693481916199099,
+      "learning_rate": 1.8031331145953047e-06,
+      "loss": 0.5181,
+      "step": 1404
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 1.01144312594075,
+      "learning_rate": 1.7959793279420507e-06,
+      "loss": 0.4639,
+      "step": 1405
+    },
+    {
+      "epoch": 2.2496,
+      "grad_norm": 1.1390245892039497,
+      "learning_rate": 1.7888366525859968e-06,
+      "loss": 0.726,
+      "step": 1406
+    },
+    {
+      "epoch": 2.2512,
+      "grad_norm": 0.9627899078520802,
+      "learning_rate": 1.781705113297396e-06,
+      "loss": 0.5431,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 1.069626037532042,
+      "learning_rate": 1.7745847348078742e-06,
+      "loss": 0.4984,
+      "step": 1408
+    },
+    {
+      "epoch": 2.2544,
+      "grad_norm": 0.8490098135377564,
+      "learning_rate": 1.7674755418103578e-06,
+      "loss": 0.4637,
+      "step": 1409
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.8947559502653272,
+      "learning_rate": 1.7603775589589821e-06,
+      "loss": 0.4916,
+      "step": 1410
+    },
+    {
+      "epoch": 2.2576,
+      "grad_norm": 0.8675150620751441,
+      "learning_rate": 1.7532908108690038e-06,
+      "loss": 0.5049,
+      "step": 1411
+    },
+    {
+      "epoch": 2.2592,
+      "grad_norm": 0.8463471938574573,
+      "learning_rate": 1.7462153221167222e-06,
+      "loss": 0.464,
+      "step": 1412
+    },
+    {
+      "epoch": 2.2608,
+      "grad_norm": 0.9755081321432716,
+      "learning_rate": 1.7391511172393849e-06,
+      "loss": 0.4378,
+      "step": 1413
+    },
+    {
+      "epoch": 2.2624,
+      "grad_norm": 0.8960837478164718,
+      "learning_rate": 1.7320982207351128e-06,
+      "loss": 0.4434,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 0.8639340206206791,
+      "learning_rate": 1.7250566570628103e-06,
+      "loss": 0.4831,
+      "step": 1415
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.9774904991221383,
+      "learning_rate": 1.7180264506420746e-06,
+      "loss": 0.5418,
+      "step": 1416
+    },
+    {
+      "epoch": 2.2672,
+      "grad_norm": 1.0198119590905121,
+      "learning_rate": 1.7110076258531244e-06,
+      "loss": 0.5515,
+      "step": 1417
+    },
+    {
+      "epoch": 2.2688,
+      "grad_norm": 1.0248449753921995,
+      "learning_rate": 1.7040002070367006e-06,
+      "loss": 0.7002,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2704,
+      "grad_norm": 0.9181124434613644,
+      "learning_rate": 1.6970042184939943e-06,
+      "loss": 0.5167,
+      "step": 1419
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 0.843391244739533,
+      "learning_rate": 1.6900196844865575e-06,
+      "loss": 0.3305,
+      "step": 1420
+    },
+    {
+      "epoch": 2.2736,
+      "grad_norm": 1.0176184370574133,
+      "learning_rate": 1.683046629236213e-06,
+      "loss": 0.5698,
+      "step": 1421
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.7801820069544,
+      "learning_rate": 1.6760850769249837e-06,
+      "loss": 0.3833,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2768,
+      "grad_norm": 0.9413475260075757,
+      "learning_rate": 1.669135051694994e-06,
+      "loss": 0.5807,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.8650839944803358,
+      "learning_rate": 1.662196577648398e-06,
+      "loss": 0.4929,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.7284311750394808,
+      "learning_rate": 1.6552696788472921e-06,
+      "loss": 0.2866,
+      "step": 1425
+    },
+    {
+      "epoch": 2.2816,
+      "grad_norm": 0.7814361283281405,
+      "learning_rate": 1.6483543793136247e-06,
+      "loss": 0.4675,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2832,
+      "grad_norm": 0.8677573887775226,
+      "learning_rate": 1.6414507030291249e-06,
+      "loss": 0.2955,
+      "step": 1427
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.988728415107249,
+      "learning_rate": 1.6345586739352105e-06,
+      "loss": 0.3697,
+      "step": 1428
+    },
+    {
+      "epoch": 2.2864,
+      "grad_norm": 0.9980071551471438,
+      "learning_rate": 1.6276783159329095e-06,
+      "loss": 0.4759,
+      "step": 1429
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 0.8668333184579307,
+      "learning_rate": 1.6208096528827717e-06,
+      "loss": 0.4137,
+      "step": 1430
+    },
+    {
+      "epoch": 2.2896,
+      "grad_norm": 1.0179803322756729,
+      "learning_rate": 1.6139527086047929e-06,
+      "loss": 0.5741,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 1.074098748460684,
+      "learning_rate": 1.6071075068783303e-06,
+      "loss": 0.7954,
+      "step": 1432
+    },
+    {
+      "epoch": 2.2928,
+      "grad_norm": 0.8968905402179511,
+      "learning_rate": 1.600274071442014e-06,
+      "loss": 0.4845,
+      "step": 1433
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.9744662666012526,
+      "learning_rate": 1.5934524259936757e-06,
+      "loss": 0.508,
+      "step": 1434
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 0.8743790885453947,
+      "learning_rate": 1.5866425941902524e-06,
+      "loss": 0.4649,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2976,
+      "grad_norm": 0.9383120061816762,
+      "learning_rate": 1.5798445996477219e-06,
+      "loss": 0.5626,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2992,
+      "grad_norm": 0.8798660620549341,
+      "learning_rate": 1.573058465941002e-06,
+      "loss": 0.4186,
+      "step": 1437
+    },
+    {
+      "epoch": 2.3008,
+      "grad_norm": 0.8984768664510973,
+      "learning_rate": 1.5662842166038844e-06,
+      "loss": 0.531,
+      "step": 1438
+    },
+    {
+      "epoch": 2.3024,
+      "grad_norm": 1.055502814148748,
+      "learning_rate": 1.5595218751289465e-06,
+      "loss": 0.5181,
+      "step": 1439
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 1.0372758022988147,
+      "learning_rate": 1.5527714649674641e-06,
+      "loss": 0.3988,
+      "step": 1440
+    },
+    {
+      "epoch": 2.3056,
+      "grad_norm": 0.9845550301675416,
+      "learning_rate": 1.5460330095293447e-06,
+      "loss": 0.5392,
+      "step": 1441
+    },
+    {
+      "epoch": 2.3072,
+      "grad_norm": 1.1404384837233654,
+      "learning_rate": 1.5393065321830292e-06,
+      "loss": 0.7446,
+      "step": 1442
+    },
+    {
+      "epoch": 2.3088,
+      "grad_norm": 1.0661318826980928,
+      "learning_rate": 1.5325920562554259e-06,
+      "loss": 0.5452,
+      "step": 1443
+    },
+    {
+      "epoch": 2.3104,
+      "grad_norm": 0.9794322474791779,
+      "learning_rate": 1.5258896050318217e-06,
+      "loss": 0.6498,
+      "step": 1444
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 0.8638692312751757,
+      "learning_rate": 1.5191992017557994e-06,
+      "loss": 0.4895,
+      "step": 1445
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.8168962693699611,
+      "learning_rate": 1.512520869629165e-06,
+      "loss": 0.436,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3152,
+      "grad_norm": 0.8752909244562628,
+      "learning_rate": 1.5058546318118583e-06,
+      "loss": 0.5138,
+      "step": 1447
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.8294668197064706,
+      "learning_rate": 1.4992005114218805e-06,
+      "loss": 0.4432,
+      "step": 1448
+    },
+    {
+      "epoch": 2.3184,
+      "grad_norm": 0.8315647849304361,
+      "learning_rate": 1.4925585315352108e-06,
+      "loss": 0.4334,
+      "step": 1449
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.9879965507006222,
+      "learning_rate": 1.485928715185721e-06,
+      "loss": 0.5348,
+      "step": 1450
+    },
+    {
+      "epoch": 2.3216,
+      "grad_norm": 0.9830872092995894,
+      "learning_rate": 1.4793110853651077e-06,
+      "loss": 0.5718,
+      "step": 1451
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 1.0264402552782776,
+      "learning_rate": 1.472705665022799e-06,
+      "loss": 0.6643,
+      "step": 1452
+    },
+    {
+      "epoch": 2.3247999999999998,
+      "grad_norm": 0.7763377515846339,
+      "learning_rate": 1.4661124770658857e-06,
+      "loss": 0.3764,
+      "step": 1453
+    },
+    {
+      "epoch": 2.3264,
+      "grad_norm": 0.7781477366631595,
+      "learning_rate": 1.459531544359038e-06,
+      "loss": 0.442,
+      "step": 1454
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 0.8469686935626068,
+      "learning_rate": 1.4529628897244214e-06,
+      "loss": 0.4833,
+      "step": 1455
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 1.0070940011655571,
+      "learning_rate": 1.4464065359416274e-06,
+      "loss": 0.6041,
+      "step": 1456
+    },
+    {
+      "epoch": 2.3312,
+      "grad_norm": 0.9647632584518745,
+      "learning_rate": 1.4398625057475845e-06,
+      "loss": 0.6421,
+      "step": 1457
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 1.0280703858494444,
+      "learning_rate": 1.4333308218364861e-06,
+      "loss": 0.6827,
+      "step": 1458
+    },
+    {
+      "epoch": 2.3344,
+      "grad_norm": 0.9011826247246635,
+      "learning_rate": 1.4268115068597122e-06,
+      "loss": 0.4688,
+      "step": 1459
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 1.0505583423555425,
+      "learning_rate": 1.4203045834257418e-06,
+      "loss": 0.5182,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3376,
+      "grad_norm": 0.8863923270697123,
+      "learning_rate": 1.4138100741000888e-06,
+      "loss": 0.5647,
+      "step": 1461
+    },
+    {
+      "epoch": 2.3392,
+      "grad_norm": 0.9505037684676528,
+      "learning_rate": 1.4073280014052077e-06,
+      "loss": 0.5377,
+      "step": 1462
+    },
+    {
+      "epoch": 2.3407999999999998,
+      "grad_norm": 0.931224401502142,
+      "learning_rate": 1.4008583878204297e-06,
+      "loss": 0.4589,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 0.7971037678282673,
+      "learning_rate": 1.3944012557818793e-06,
+      "loss": 0.4005,
+      "step": 1464
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 1.0291454845860915,
+      "learning_rate": 1.3879566276823896e-06,
+      "loss": 0.6464,
+      "step": 1465
+    },
+    {
+      "epoch": 2.3456,
+      "grad_norm": 0.9351743241184531,
+      "learning_rate": 1.3815245258714393e-06,
+      "loss": 0.525,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3472,
+      "grad_norm": 0.9228173981569134,
+      "learning_rate": 1.3751049726550587e-06,
+      "loss": 0.3739,
+      "step": 1467
+    },
+    {
+      "epoch": 2.3487999999999998,
+      "grad_norm": 0.7941430580623614,
+      "learning_rate": 1.368697990295766e-06,
+      "loss": 0.3716,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3504,
+      "grad_norm": 0.8233785223510821,
+      "learning_rate": 1.3623036010124845e-06,
+      "loss": 0.4428,
+      "step": 1469
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.8634769007152763,
+      "learning_rate": 1.3559218269804624e-06,
+      "loss": 0.3871,
+      "step": 1470
+    },
+    {
+      "epoch": 2.3536,
+      "grad_norm": 0.9427682415927455,
+      "learning_rate": 1.3495526903312029e-06,
+      "loss": 0.5425,
+      "step": 1471
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 1.0349456210820542,
+      "learning_rate": 1.3431962131523796e-06,
+      "loss": 0.4696,
+      "step": 1472
+    },
+    {
+      "epoch": 2.3568,
+      "grad_norm": 1.0911039835402547,
+      "learning_rate": 1.3368524174877679e-06,
+      "loss": 0.5771,
+      "step": 1473
+    },
+    {
+      "epoch": 2.3584,
+      "grad_norm": 0.7968514416140384,
+      "learning_rate": 1.330521325337164e-06,
+      "loss": 0.3959,
+      "step": 1474
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.8845851692959773,
+      "learning_rate": 1.3242029586563054e-06,
+      "loss": 0.4718,
+      "step": 1475
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.7977987843392154,
+      "learning_rate": 1.3178973393568055e-06,
+      "loss": 0.3798,
+      "step": 1476
+    },
+    {
+      "epoch": 2.3632,
+      "grad_norm": 1.032462837463176,
+      "learning_rate": 1.3116044893060637e-06,
+      "loss": 0.4821,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3648,
+      "grad_norm": 0.9211348321427932,
+      "learning_rate": 1.3053244303272022e-06,
+      "loss": 0.5409,
+      "step": 1478
+    },
+    {
+      "epoch": 2.3664,
+      "grad_norm": 1.1259103399975388,
+      "learning_rate": 1.2990571841989796e-06,
+      "loss": 0.6565,
+      "step": 1479
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.8419940597978801,
+      "learning_rate": 1.2928027726557257e-06,
+      "loss": 0.5082,
+      "step": 1480
+    },
+    {
+      "epoch": 2.3696,
+      "grad_norm": 0.9036754191302891,
+      "learning_rate": 1.2865612173872577e-06,
+      "loss": 0.4981,
+      "step": 1481
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.9593716393505523,
+      "learning_rate": 1.2803325400388095e-06,
+      "loss": 0.562,
+      "step": 1482
+    },
+    {
+      "epoch": 2.3728,
+      "grad_norm": 1.050087787959598,
+      "learning_rate": 1.2741167622109557e-06,
+      "loss": 0.6101,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3744,
+      "grad_norm": 0.7985720428533846,
+      "learning_rate": 1.2679139054595335e-06,
+      "loss": 0.3355,
+      "step": 1484
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 0.9576163964466468,
+      "learning_rate": 1.261723991295576e-06,
+      "loss": 0.5638,
+      "step": 1485
+    },
+    {
+      "epoch": 2.3776,
+      "grad_norm": 0.6272081710935784,
+      "learning_rate": 1.2555470411852262e-06,
+      "loss": 0.2753,
+      "step": 1486
+    },
+    {
+      "epoch": 2.3792,
+      "grad_norm": 1.0350813435280268,
+      "learning_rate": 1.2493830765496724e-06,
+      "loss": 0.5176,
+      "step": 1487
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 1.0808997099057434,
+      "learning_rate": 1.2432321187650726e-06,
+      "loss": 0.643,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3824,
+      "grad_norm": 0.8941009523873167,
+      "learning_rate": 1.237094189162471e-06,
+      "loss": 0.4858,
+      "step": 1489
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 1.1017321220132814,
+      "learning_rate": 1.2309693090277392e-06,
+      "loss": 0.611,
+      "step": 1490
+    },
+    {
+      "epoch": 2.3856,
+      "grad_norm": 0.8932461876462147,
+      "learning_rate": 1.2248574996014872e-06,
+      "loss": 0.5659,
+      "step": 1491
+    },
+    {
+      "epoch": 2.3872,
+      "grad_norm": 0.9925122957434063,
+      "learning_rate": 1.218758782079001e-06,
+      "loss": 0.4996,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3888,
+      "grad_norm": 0.856964661362695,
+      "learning_rate": 1.2126731776101657e-06,
+      "loss": 0.4799,
+      "step": 1493
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.9607934892127727,
+      "learning_rate": 1.2066007072993856e-06,
+      "loss": 0.6064,
+      "step": 1494
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 0.9857229283372475,
+      "learning_rate": 1.2005413922055247e-06,
+      "loss": 0.5506,
+      "step": 1495
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 1.3128733793715772,
+      "learning_rate": 1.194495253341818e-06,
+      "loss": 0.3863,
+      "step": 1496
+    },
+    {
+      "epoch": 2.3952,
+      "grad_norm": 0.9698894912499546,
+      "learning_rate": 1.1884623116758121e-06,
+      "loss": 0.4183,
+      "step": 1497
+    },
+    {
+      "epoch": 2.3968,
+      "grad_norm": 0.9358455370415274,
+      "learning_rate": 1.1824425881292846e-06,
+      "loss": 0.5372,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3984,
+      "grad_norm": 0.8346624843186394,
+      "learning_rate": 1.1764361035781718e-06,
+      "loss": 0.5429,
+      "step": 1499
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7409267284022,
+      "learning_rate": 1.170442878852503e-06,
+      "loss": 0.3651,
+      "step": 1500
+    },
+    {
+      "epoch": 2.4016,
+      "grad_norm": 0.8558318301902453,
+      "learning_rate": 1.1644629347363173e-06,
+      "loss": 0.5684,
+      "step": 1501
+    },
+    {
+      "epoch": 2.4032,
+      "grad_norm": 0.9717717372745454,
+      "learning_rate": 1.1584962919676024e-06,
+      "loss": 0.5896,
+      "step": 1502
+    },
+    {
+      "epoch": 2.4048,
+      "grad_norm": 0.9175508481566372,
+      "learning_rate": 1.1525429712382175e-06,
+      "loss": 0.4913,
+      "step": 1503
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.9602325597251903,
+      "learning_rate": 1.1466029931938182e-06,
+      "loss": 0.5029,
+      "step": 1504
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 0.9363830209724805,
+      "learning_rate": 1.1406763784337948e-06,
+      "loss": 0.657,
+      "step": 1505
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.9086781042516432,
+      "learning_rate": 1.1347631475111882e-06,
+      "loss": 0.6016,
+      "step": 1506
+    },
+    {
+      "epoch": 2.4112,
+      "grad_norm": 0.9192220642753022,
+      "learning_rate": 1.1288633209326288e-06,
+      "loss": 0.541,
+      "step": 1507
+    },
+    {
+      "epoch": 2.4128,
+      "grad_norm": 0.9886734868994789,
+      "learning_rate": 1.122976919158264e-06,
+      "loss": 0.548,
+      "step": 1508
+    },
+    {
+      "epoch": 2.4144,
+      "grad_norm": 1.0886591503792469,
+      "learning_rate": 1.1171039626016789e-06,
+      "loss": 0.7095,
+      "step": 1509
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 1.0004153226831678,
+      "learning_rate": 1.1112444716298381e-06,
+      "loss": 0.6341,
+      "step": 1510
+    },
+    {
+      "epoch": 2.4176,
+      "grad_norm": 0.9269518410932112,
+      "learning_rate": 1.1053984665630025e-06,
+      "loss": 0.5398,
+      "step": 1511
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 1.0159500162668893,
+      "learning_rate": 1.0995659676746706e-06,
+      "loss": 0.7151,
+      "step": 1512
+    },
+    {
+      "epoch": 2.4208,
+      "grad_norm": 0.9049903122726919,
+      "learning_rate": 1.093746995191497e-06,
+      "loss": 0.6204,
+      "step": 1513
+    },
+    {
+      "epoch": 2.4224,
+      "grad_norm": 0.8783189023130342,
+      "learning_rate": 1.0879415692932328e-06,
+      "loss": 0.4715,
+      "step": 1514
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 0.96999410358796,
+      "learning_rate": 1.0821497101126487e-06,
+      "loss": 0.5089,
+      "step": 1515
+    },
+    {
+      "epoch": 2.4256,
+      "grad_norm": 0.9896969889815371,
+      "learning_rate": 1.076371437735465e-06,
+      "loss": 0.6389,
+      "step": 1516
+    },
+    {
+      "epoch": 2.4272,
+      "grad_norm": 0.9937005001862865,
+      "learning_rate": 1.0706067722002877e-06,
+      "loss": 0.3535,
+      "step": 1517
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.9283500157098227,
+      "learning_rate": 1.064855733498531e-06,
+      "loss": 0.4832,
+      "step": 1518
+    },
+    {
+      "epoch": 2.4304,
+      "grad_norm": 0.9991008224045489,
+      "learning_rate": 1.0591183415743562e-06,
+      "loss": 0.6736,
+      "step": 1519
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.8675201089133293,
+      "learning_rate": 1.0533946163245984e-06,
+      "loss": 0.4446,
+      "step": 1520
+    },
+    {
+      "epoch": 2.4336,
+      "grad_norm": 0.850469476842336,
+      "learning_rate": 1.047684577598694e-06,
+      "loss": 0.2885,
+      "step": 1521
+    },
+    {
+      "epoch": 2.4352,
+      "grad_norm": 1.1584340915993605,
+      "learning_rate": 1.0419882451986197e-06,
+      "loss": 0.6674,
+      "step": 1522
+    },
+    {
+      "epoch": 2.4368,
+      "grad_norm": 0.8894318984533174,
+      "learning_rate": 1.0363056388788162e-06,
+      "loss": 0.4195,
+      "step": 1523
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.9408991920127242,
+      "learning_rate": 1.0306367783461258e-06,
+      "loss": 0.605,
+      "step": 1524
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7745548391269444,
+      "learning_rate": 1.024981683259723e-06,
+      "loss": 0.3808,
+      "step": 1525
+    },
+    {
+      "epoch": 2.4416,
+      "grad_norm": 0.9630487567304588,
+      "learning_rate": 1.0193403732310392e-06,
+      "loss": 0.5986,
+      "step": 1526
+    },
+    {
+      "epoch": 2.4432,
+      "grad_norm": 0.8521058804123267,
+      "learning_rate": 1.0137128678237062e-06,
+      "loss": 0.378,
+      "step": 1527
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.8999044113182457,
+      "learning_rate": 1.0080991865534773e-06,
+      "loss": 0.5278,
+      "step": 1528
+    },
+    {
+      "epoch": 2.4464,
+      "grad_norm": 0.7952861585603319,
+      "learning_rate": 1.002499348888169e-06,
+      "loss": 0.4236,
+      "step": 1529
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 1.0261402005048572,
+      "learning_rate": 9.969133742475883e-07,
+      "loss": 0.4224,
+      "step": 1530
+    },
+    {
+      "epoch": 2.4496,
+      "grad_norm": 0.9889316838050909,
+      "learning_rate": 9.913412820034629e-07,
+      "loss": 0.574,
+      "step": 1531
+    },
+    {
+      "epoch": 2.4512,
+      "grad_norm": 0.9932429518454472,
+      "learning_rate": 9.857830914793827e-07,
+      "loss": 0.5665,
+      "step": 1532
+    },
+    {
+      "epoch": 2.4528,
+      "grad_norm": 0.9307201521631674,
+      "learning_rate": 9.802388219507215e-07,
+      "loss": 0.5696,
+      "step": 1533
+    },
+    {
+      "epoch": 2.4544,
+      "grad_norm": 1.157043761531579,
+      "learning_rate": 9.747084926445839e-07,
+      "loss": 0.8006,
+      "step": 1534
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 1.106887132673668,
+      "learning_rate": 9.691921227397227e-07,
+      "loss": 0.6043,
+      "step": 1535
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.8563532938121887,
+      "learning_rate": 9.63689731366486e-07,
+      "loss": 0.48,
+      "step": 1536
+    },
+    {
+      "epoch": 2.4592,
+      "grad_norm": 0.8815511850441942,
+      "learning_rate": 9.58201337606745e-07,
+      "loss": 0.3275,
+      "step": 1537
+    },
+    {
+      "epoch": 2.4608,
+      "grad_norm": 1.0182195024050973,
+      "learning_rate": 9.527269604938249e-07,
+      "loss": 0.6277,
+      "step": 1538
+    },
+    {
+      "epoch": 2.4624,
+      "grad_norm": 0.903571405066688,
+      "learning_rate": 9.472666190124457e-07,
+      "loss": 0.4873,
+      "step": 1539
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 0.7410404322567221,
+      "learning_rate": 9.418203320986502e-07,
+      "loss": 0.4488,
+      "step": 1540
+    },
+    {
+      "epoch": 2.4656000000000002,
+      "grad_norm": 1.2921935152056336,
+      "learning_rate": 9.363881186397434e-07,
+      "loss": 0.5158,
+      "step": 1541
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.9259953947882726,
+      "learning_rate": 9.309699974742243e-07,
+      "loss": 0.5751,
+      "step": 1542
+    },
+    {
+      "epoch": 2.4688,
+      "grad_norm": 0.8295853532147452,
+      "learning_rate": 9.255659873917183e-07,
+      "loss": 0.4862,
+      "step": 1543
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 1.1125411360401074,
+      "learning_rate": 9.201761071329196e-07,
+      "loss": 0.676,
+      "step": 1544
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 1.055931225116287,
+      "learning_rate": 9.148003753895146e-07,
+      "loss": 0.5084,
+      "step": 1545
+    },
+    {
+      "epoch": 2.4736000000000002,
+      "grad_norm": 1.0878368525055406,
+      "learning_rate": 9.094388108041302e-07,
+      "loss": 0.6544,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4752,
+      "grad_norm": 0.8039901790854609,
+      "learning_rate": 9.040914319702598e-07,
+      "loss": 0.3459,
+      "step": 1547
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.9684094335087292,
+      "learning_rate": 8.987582574321996e-07,
+      "loss": 0.4744,
+      "step": 1548
+    },
+    {
+      "epoch": 2.4784,
+      "grad_norm": 1.0787715686927517,
+      "learning_rate": 8.934393056849921e-07,
+      "loss": 0.7146,
+      "step": 1549
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.106324480672612,
+      "learning_rate": 8.881345951743486e-07,
+      "loss": 0.6958,
+      "step": 1550
+    },
+    {
+      "epoch": 2.4816,
+      "grad_norm": 0.8396125451035065,
+      "learning_rate": 8.828441442966013e-07,
+      "loss": 0.4374,
+      "step": 1551
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.7816353651829983,
+      "learning_rate": 8.775679713986235e-07,
+      "loss": 0.4796,
+      "step": 1552
+    },
+    {
+      "epoch": 2.4848,
+      "grad_norm": 0.9704276003053988,
+      "learning_rate": 8.723060947777778e-07,
+      "loss": 0.5111,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.8795115315809064,
+      "learning_rate": 8.670585326818493e-07,
+      "loss": 0.4143,
+      "step": 1554
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 0.9493986391982322,
+      "learning_rate": 8.618253033089768e-07,
+      "loss": 0.5722,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4896,
+      "grad_norm": 0.9417247177160813,
+      "learning_rate": 8.566064248076001e-07,
+      "loss": 0.576,
+      "step": 1556
+    },
+    {
+      "epoch": 2.4912,
+      "grad_norm": 0.9731896737599867,
+      "learning_rate": 8.514019152763852e-07,
+      "loss": 0.686,
+      "step": 1557
+    },
+    {
+      "epoch": 2.4928,
+      "grad_norm": 0.833465156765236,
+      "learning_rate": 8.462117927641733e-07,
+      "loss": 0.4584,
+      "step": 1558
+    },
+    {
+      "epoch": 2.4944,
+      "grad_norm": 1.021932582279391,
+      "learning_rate": 8.410360752699099e-07,
+      "loss": 0.5372,
+      "step": 1559
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.987746242846434,
+      "learning_rate": 8.358747807425827e-07,
+      "loss": 0.5807,
+      "step": 1560
+    },
+    {
+      "epoch": 2.4976,
+      "grad_norm": 0.8571418458806283,
+      "learning_rate": 8.307279270811675e-07,
+      "loss": 0.4619,
+      "step": 1561
+    },
+    {
+      "epoch": 2.4992,
+      "grad_norm": 1.0132588448661046,
+      "learning_rate": 8.255955321345533e-07,
+      "loss": 0.4804,
+      "step": 1562
+    },
+    {
+      "epoch": 2.5008,
+      "grad_norm": 0.6658333060229576,
+      "learning_rate": 8.20477613701493e-07,
+      "loss": 0.3223,
+      "step": 1563
+    },
+    {
+      "epoch": 2.5023999999999997,
+      "grad_norm": 0.9729007017782386,
+      "learning_rate": 8.153741895305351e-07,
+      "loss": 0.5263,
+      "step": 1564
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 0.9082295959977547,
+      "learning_rate": 8.102852773199588e-07,
+      "loss": 0.4918,
+      "step": 1565
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.8262162530638266,
+      "learning_rate": 8.052108947177234e-07,
+      "loss": 0.4031,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5072,
+      "grad_norm": 0.7669844563759162,
+      "learning_rate": 8.001510593213946e-07,
+      "loss": 0.4422,
+      "step": 1567
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.9280154460923005,
+      "learning_rate": 7.951057886780939e-07,
+      "loss": 0.3876,
+      "step": 1568
+    },
+    {
+      "epoch": 2.5103999999999997,
+      "grad_norm": 0.9174429520819104,
+      "learning_rate": 7.900751002844326e-07,
+      "loss": 0.5939,
+      "step": 1569
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 0.8309234305849474,
+      "learning_rate": 7.850590115864481e-07,
+      "loss": 0.5157,
+      "step": 1570
+    },
+    {
+      "epoch": 2.5136,
+      "grad_norm": 0.98566309716522,
+      "learning_rate": 7.80057539979554e-07,
+      "loss": 0.5394,
+      "step": 1571
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 1.0097208241476237,
+      "learning_rate": 7.750707028084653e-07,
+      "loss": 0.5979,
+      "step": 1572
+    },
+    {
+      "epoch": 2.5168,
+      "grad_norm": 1.0951443744402032,
+      "learning_rate": 7.70098517367151e-07,
+      "loss": 0.8492,
+      "step": 1573
+    },
+    {
+      "epoch": 2.5183999999999997,
+      "grad_norm": 0.897419409790903,
+      "learning_rate": 7.651410008987698e-07,
+      "loss": 0.472,
+      "step": 1574
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.8876981218990583,
+      "learning_rate": 7.601981705956041e-07,
+      "loss": 0.5103,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.9079531631030493,
+      "learning_rate": 7.552700435990123e-07,
+      "loss": 0.561,
+      "step": 1576
+    },
+    {
+      "epoch": 2.5232,
+      "grad_norm": 0.7017054799714203,
+      "learning_rate": 7.503566369993564e-07,
+      "loss": 0.2813,
+      "step": 1577
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.892673973549526,
+      "learning_rate": 7.454579678359547e-07,
+      "loss": 0.5453,
+      "step": 1578
+    },
+    {
+      "epoch": 2.5263999999999998,
+      "grad_norm": 0.917987461622968,
+      "learning_rate": 7.405740530970157e-07,
+      "loss": 0.5609,
+      "step": 1579
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 0.867083338819329,
+      "learning_rate": 7.357049097195773e-07,
+      "loss": 0.3466,
+      "step": 1580
+    },
+    {
+      "epoch": 2.5296,
+      "grad_norm": 1.058392826757329,
+      "learning_rate": 7.308505545894567e-07,
+      "loss": 0.6278,
+      "step": 1581
+    },
+    {
+      "epoch": 2.5312,
+      "grad_norm": 1.0141113237486776,
+      "learning_rate": 7.260110045411816e-07,
+      "loss": 0.5813,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5328,
+      "grad_norm": 0.7162891327285343,
+      "learning_rate": 7.211862763579414e-07,
+      "loss": 0.311,
+      "step": 1583
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.8758030852091689,
+      "learning_rate": 7.163763867715218e-07,
+      "loss": 0.4811,
+      "step": 1584
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 0.9897671676437568,
+      "learning_rate": 7.115813524622489e-07,
+      "loss": 0.672,
+      "step": 1585
+    },
+    {
+      "epoch": 2.5376,
+      "grad_norm": 0.9949710524025478,
+      "learning_rate": 7.068011900589333e-07,
+      "loss": 0.4206,
+      "step": 1586
+    },
+    {
+      "epoch": 2.5392,
+      "grad_norm": 0.8837911187184745,
+      "learning_rate": 7.020359161388108e-07,
+      "loss": 0.4316,
+      "step": 1587
+    },
+    {
+      "epoch": 2.5408,
+      "grad_norm": 0.5879184112503804,
+      "learning_rate": 6.972855472274853e-07,
+      "loss": 0.2481,
+      "step": 1588
+    },
+    {
+      "epoch": 2.5423999999999998,
+      "grad_norm": 0.9759203147374813,
+      "learning_rate": 6.925500997988694e-07,
+      "loss": 0.5992,
+      "step": 1589
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.9498676076602623,
+      "learning_rate": 6.87829590275132e-07,
+      "loss": 0.4977,
+      "step": 1590
+    },
+    {
+      "epoch": 2.5456,
+      "grad_norm": 1.029532618020753,
+      "learning_rate": 6.83124035026635e-07,
+      "loss": 0.5373,
+      "step": 1591
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.668204761482299,
+      "learning_rate": 6.784334503718826e-07,
+      "loss": 0.3394,
+      "step": 1592
+    },
+    {
+      "epoch": 2.5488,
+      "grad_norm": 0.8695594939812411,
+      "learning_rate": 6.737578525774636e-07,
+      "loss": 0.4441,
+      "step": 1593
+    },
+    {
+      "epoch": 2.5504,
+      "grad_norm": 0.9392349325484638,
+      "learning_rate": 6.690972578579886e-07,
+      "loss": 0.59,
+      "step": 1594
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 0.9078163588148144,
+      "learning_rate": 6.644516823760439e-07,
+      "loss": 0.52,
+      "step": 1595
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.8039329651478745,
+      "learning_rate": 6.598211422421258e-07,
+      "loss": 0.4519,
+      "step": 1596
+    },
+    {
+      "epoch": 2.5552,
+      "grad_norm": 0.9627544122110097,
+      "learning_rate": 6.552056535145917e-07,
+      "loss": 0.5289,
+      "step": 1597
+    },
+    {
+      "epoch": 2.5568,
+      "grad_norm": 0.9840575022262902,
+      "learning_rate": 6.506052321996037e-07,
+      "loss": 0.5237,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5584,
+      "grad_norm": 0.9787018167214986,
+      "learning_rate": 6.46019894251066e-07,
+      "loss": 0.5315,
+      "step": 1599
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.0493059715927147,
+      "learning_rate": 6.414496555705802e-07,
+      "loss": 0.5765,
+      "step": 1600
+    },
+    {
+      "epoch": 2.5616,
+      "grad_norm": 0.8969651754034742,
+      "learning_rate": 6.368945320073799e-07,
+      "loss": 0.5299,
+      "step": 1601
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.945206524039645,
+      "learning_rate": 6.323545393582847e-07,
+      "loss": 0.5838,
+      "step": 1602
+    },
+    {
+      "epoch": 2.5648,
+      "grad_norm": 1.0258998212558372,
+      "learning_rate": 6.278296933676414e-07,
+      "loss": 0.4789,
+      "step": 1603
+    },
+    {
+      "epoch": 2.5664,
+      "grad_norm": 0.9473643513353971,
+      "learning_rate": 6.233200097272646e-07,
+      "loss": 0.5551,
+      "step": 1604
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 0.7774430184665326,
+      "learning_rate": 6.188255040763929e-07,
+      "loss": 0.3734,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5696,
+      "grad_norm": 0.9445584853829107,
+      "learning_rate": 6.143461920016247e-07,
+      "loss": 0.5883,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5712,
+      "grad_norm": 0.9296917836396839,
+      "learning_rate": 6.098820890368696e-07,
+      "loss": 0.5147,
+      "step": 1607
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.9148586062615583,
+      "learning_rate": 6.054332106632943e-07,
+      "loss": 0.5732,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5744,
+      "grad_norm": 1.0012412849625194,
+      "learning_rate": 6.009995723092655e-07,
+      "loss": 0.5664,
+      "step": 1609
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 1.0987958419551875,
+      "learning_rate": 5.965811893503015e-07,
+      "loss": 0.6781,
+      "step": 1610
+    },
+    {
+      "epoch": 2.5776,
+      "grad_norm": 0.7690880889200417,
+      "learning_rate": 5.921780771090124e-07,
+      "loss": 0.3247,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5792,
+      "grad_norm": 0.939748803044778,
+      "learning_rate": 5.877902508550542e-07,
+      "loss": 0.4916,
+      "step": 1612
+    },
+    {
+      "epoch": 2.5808,
+      "grad_norm": 1.0847911262550938,
+      "learning_rate": 5.834177258050711e-07,
+      "loss": 0.718,
+      "step": 1613
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 1.070434943448536,
+      "learning_rate": 5.790605171226421e-07,
+      "loss": 0.4535,
+      "step": 1614
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 0.9753736580159518,
+      "learning_rate": 5.747186399182336e-07,
+      "loss": 0.433,
+      "step": 1615
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.9386288978432279,
+      "learning_rate": 5.703921092491393e-07,
+      "loss": 0.5085,
+      "step": 1616
+    },
+    {
+      "epoch": 2.5872,
+      "grad_norm": 0.8610325004451929,
+      "learning_rate": 5.660809401194362e-07,
+      "loss": 0.4918,
+      "step": 1617
+    },
+    {
+      "epoch": 2.5888,
+      "grad_norm": 0.9401352428232486,
+      "learning_rate": 5.617851474799285e-07,
+      "loss": 0.569,
+      "step": 1618
+    },
+    {
+      "epoch": 2.5904,
+      "grad_norm": 1.0696131085570977,
+      "learning_rate": 5.575047462280919e-07,
+      "loss": 0.6889,
+      "step": 1619
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.8732391087759279,
+      "learning_rate": 5.532397512080306e-07,
+      "loss": 0.4544,
+      "step": 1620
+    },
+    {
+      "epoch": 2.5936,
+      "grad_norm": 1.0847734004509193,
+      "learning_rate": 5.489901772104178e-07,
+      "loss": 0.5871,
+      "step": 1621
+    },
+    {
+      "epoch": 2.5952,
+      "grad_norm": 1.025147023137289,
+      "learning_rate": 5.447560389724499e-07,
+      "loss": 0.5823,
+      "step": 1622
+    },
+    {
+      "epoch": 2.5968,
+      "grad_norm": 0.9376440736222197,
+      "learning_rate": 5.405373511777939e-07,
+      "loss": 0.586,
+      "step": 1623
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.751587873002503,
+      "learning_rate": 5.363341284565316e-07,
+      "loss": 0.3986,
+      "step": 1624
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.909613639993501,
+      "learning_rate": 5.321463853851189e-07,
+      "loss": 0.5104,
+      "step": 1625
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.8889224427108126,
+      "learning_rate": 5.279741364863244e-07,
+      "loss": 0.5365,
+      "step": 1626
+    },
+    {
+      "epoch": 2.6032,
+      "grad_norm": 0.7634680142042353,
+      "learning_rate": 5.238173962291881e-07,
+      "loss": 0.4034,
+      "step": 1627
+    },
+    {
+      "epoch": 2.6048,
+      "grad_norm": 0.956085254453705,
+      "learning_rate": 5.196761790289639e-07,
+      "loss": 0.593,
+      "step": 1628
+    },
+    {
+      "epoch": 2.6064,
+      "grad_norm": 0.9012307126570898,
+      "learning_rate": 5.155504992470751e-07,
+      "loss": 0.4786,
+      "step": 1629
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.8284727307964317,
+      "learning_rate": 5.114403711910631e-07,
+      "loss": 0.4324,
+      "step": 1630
+    },
+    {
+      "epoch": 2.6096,
+      "grad_norm": 0.9106223455704877,
+      "learning_rate": 5.073458091145328e-07,
+      "loss": 0.4896,
+      "step": 1631
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.9496870308486733,
+      "learning_rate": 5.032668272171138e-07,
+      "loss": 0.4734,
+      "step": 1632
+    },
+    {
+      "epoch": 2.6128,
+      "grad_norm": 0.7895275786110693,
+      "learning_rate": 4.99203439644399e-07,
+      "loss": 0.4637,
+      "step": 1633
+    },
+    {
+      "epoch": 2.6144,
+      "grad_norm": 1.0632457095304344,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.6212,
+      "step": 1634
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 0.8155102509323091,
+      "learning_rate": 4.911235037850187e-07,
+      "loss": 0.4942,
+      "step": 1635
+    },
+    {
+      "epoch": 2.6176,
+      "grad_norm": 0.8802028987780882,
+      "learning_rate": 4.871069835189485e-07,
+      "loss": 0.5143,
+      "step": 1636
+    },
+    {
+      "epoch": 2.6192,
+      "grad_norm": 0.7785846291097843,
+      "learning_rate": 4.831061136186787e-07,
+      "loss": 0.4541,
+      "step": 1637
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.6700390055007225,
+      "learning_rate": 4.791209079589165e-07,
+      "loss": 0.3529,
+      "step": 1638
+    },
+    {
+      "epoch": 2.6224,
+      "grad_norm": 0.8686085163503766,
+      "learning_rate": 4.7515138036005157e-07,
+      "loss": 0.257,
+      "step": 1639
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 1.0339388063236605,
+      "learning_rate": 4.7119754458809727e-07,
+      "loss": 0.6319,
+      "step": 1640
+    },
+    {
+      "epoch": 2.6256,
+      "grad_norm": 0.8883200268383784,
+      "learning_rate": 4.672594143546538e-07,
+      "loss": 0.5812,
+      "step": 1641
+    },
+    {
+      "epoch": 2.6272,
+      "grad_norm": 0.6793900061951278,
+      "learning_rate": 4.6333700331685385e-07,
+      "loss": 0.3194,
+      "step": 1642
+    },
+    {
+      "epoch": 2.6288,
+      "grad_norm": 0.8686822738965762,
+      "learning_rate": 4.594303250773152e-07,
+      "loss": 0.5448,
+      "step": 1643
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 1.0549996185261559,
+      "learning_rate": 4.555393931841001e-07,
+      "loss": 0.7464,
+      "step": 1644
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 0.7102411307506447,
+      "learning_rate": 4.5166422113065877e-07,
+      "loss": 0.4068,
+      "step": 1645
+    },
+    {
+      "epoch": 2.6336,
+      "grad_norm": 1.0870681513982696,
+      "learning_rate": 4.478048223557907e-07,
+      "loss": 0.6791,
+      "step": 1646
+    },
+    {
+      "epoch": 2.6352,
+      "grad_norm": 0.8516742917262574,
+      "learning_rate": 4.439612102435942e-07,
+      "loss": 0.4488,
+      "step": 1647
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.7222784814074833,
+      "learning_rate": 4.401333981234196e-07,
+      "loss": 0.3667,
+      "step": 1648
+    },
+    {
+      "epoch": 2.6384,
+      "grad_norm": 0.9613780260833652,
+      "learning_rate": 4.3632139926982676e-07,
+      "loss": 0.582,
+      "step": 1649
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.7096611127129745,
+      "learning_rate": 4.325252269025315e-07,
+      "loss": 0.3525,
+      "step": 1650
+    },
+    {
+      "epoch": 2.6416,
+      "grad_norm": 0.8206223522864046,
+      "learning_rate": 4.287448941863692e-07,
+      "loss": 0.412,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6432,
+      "grad_norm": 0.9826387039466941,
+      "learning_rate": 4.249804142312436e-07,
+      "loss": 0.483,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6448,
+      "grad_norm": 1.1378638082145927,
+      "learning_rate": 4.2123180009207956e-07,
+      "loss": 0.8569,
+      "step": 1653
+    },
+    {
+      "epoch": 2.6464,
+      "grad_norm": 0.8451747231880498,
+      "learning_rate": 4.1749906476878486e-07,
+      "loss": 0.5357,
+      "step": 1654
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 0.837036012098896,
+      "learning_rate": 4.137822212061965e-07,
+      "loss": 0.2811,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.7085629675834364,
+      "learning_rate": 4.100812822940431e-07,
+      "loss": 0.3804,
+      "step": 1656
+    },
+    {
+      "epoch": 2.6512000000000002,
+      "grad_norm": 0.8050058514816502,
+      "learning_rate": 4.063962608668959e-07,
+      "loss": 0.4265,
+      "step": 1657
+    },
+    {
+      "epoch": 2.6528,
+      "grad_norm": 0.9864448067492304,
+      "learning_rate": 4.0272716970412516e-07,
+      "loss": 0.6459,
+      "step": 1658
+    },
+    {
+      "epoch": 2.6544,
+      "grad_norm": 0.733187994911944,
+      "learning_rate": 3.990740215298583e-07,
+      "loss": 0.3094,
+      "step": 1659
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 0.9090857610700898,
+      "learning_rate": 3.954368290129301e-07,
+      "loss": 0.5743,
+      "step": 1660
+    },
+    {
+      "epoch": 2.6576,
+      "grad_norm": 0.9526805483802787,
+      "learning_rate": 3.918156047668453e-07,
+      "loss": 0.4951,
+      "step": 1661
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.9280885690908339,
+      "learning_rate": 3.882103613497318e-07,
+      "loss": 0.4517,
+      "step": 1662
+    },
+    {
+      "epoch": 2.6608,
+      "grad_norm": 0.9812628974893594,
+      "learning_rate": 3.84621111264295e-07,
+      "loss": 0.6679,
+      "step": 1663
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 1.0611443751067686,
+      "learning_rate": 3.810478669577794e-07,
+      "loss": 0.554,
+      "step": 1664
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 1.0401308351120422,
+      "learning_rate": 3.7749064082191976e-07,
+      "loss": 0.5827,
+      "step": 1665
+    },
+    {
+      "epoch": 2.6656,
+      "grad_norm": 1.0567408640780052,
+      "learning_rate": 3.739494451929049e-07,
+      "loss": 0.5485,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6672000000000002,
+      "grad_norm": 0.7710697697456071,
+      "learning_rate": 3.7042429235132625e-07,
+      "loss": 0.3881,
+      "step": 1667
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 1.0333252902293386,
+      "learning_rate": 3.6691519452214387e-07,
+      "loss": 0.5227,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6704,
+      "grad_norm": 0.9040887004913815,
+      "learning_rate": 3.6342216387464047e-07,
+      "loss": 0.5323,
+      "step": 1669
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.8727130976646734,
+      "learning_rate": 3.5994521252237516e-07,
+      "loss": 0.4935,
+      "step": 1670
+    },
+    {
+      "epoch": 2.6736,
+      "grad_norm": 0.9491527846045139,
+      "learning_rate": 3.564843525231498e-07,
+      "loss": 0.5499,
+      "step": 1671
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.9880513577886809,
+      "learning_rate": 3.53039595878959e-07,
+      "loss": 0.6208,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6768,
+      "grad_norm": 1.0024124123974392,
+      "learning_rate": 3.496109545359544e-07,
+      "loss": 0.511,
+      "step": 1673
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 1.0418010399506832,
+      "learning_rate": 3.461984403844015e-07,
+      "loss": 0.6665,
+      "step": 1674
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7477781740862339,
+      "learning_rate": 3.42802065258635e-07,
+      "loss": 0.4415,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6816,
+      "grad_norm": 0.8084004133474733,
+      "learning_rate": 3.394218409370242e-07,
+      "loss": 0.4088,
+      "step": 1676
+    },
+    {
+      "epoch": 2.6832000000000003,
+      "grad_norm": 0.94420241995701,
+      "learning_rate": 3.360577791419256e-07,
+      "loss": 0.45,
+      "step": 1677
+    },
+    {
+      "epoch": 2.6848,
+      "grad_norm": 0.8999911747078578,
+      "learning_rate": 3.3270989153964707e-07,
+      "loss": 0.592,
+      "step": 1678
+    },
+    {
+      "epoch": 2.6864,
+      "grad_norm": 0.89087048353379,
+      "learning_rate": 3.2937818974040637e-07,
+      "loss": 0.5544,
+      "step": 1679
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.9665755557178914,
+      "learning_rate": 3.260626852982873e-07,
+      "loss": 0.4308,
+      "step": 1680
+    },
+    {
+      "epoch": 2.6896,
+      "grad_norm": 0.8125928666137122,
+      "learning_rate": 3.227633897112059e-07,
+      "loss": 0.4404,
+      "step": 1681
+    },
+    {
+      "epoch": 2.6912000000000003,
+      "grad_norm": 0.9657970839652169,
+      "learning_rate": 3.194803144208636e-07,
+      "loss": 0.541,
+      "step": 1682
+    },
+    {
+      "epoch": 2.6928,
+      "grad_norm": 0.9658870663260312,
+      "learning_rate": 3.16213470812714e-07,
+      "loss": 0.6173,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6944,
+      "grad_norm": 0.9858585683173319,
+      "learning_rate": 3.129628702159204e-07,
+      "loss": 0.7017,
+      "step": 1684
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 0.9753526322111876,
+      "learning_rate": 3.097285239033138e-07,
+      "loss": 0.5771,
+      "step": 1685
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.8203170630873676,
+      "learning_rate": 3.0651044309136016e-07,
+      "loss": 0.4961,
+      "step": 1686
+    },
+    {
+      "epoch": 2.6992000000000003,
+      "grad_norm": 1.0328298201204134,
+      "learning_rate": 3.033086389401141e-07,
+      "loss": 0.6105,
+      "step": 1687
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.7948792744432925,
+      "learning_rate": 3.0012312255318696e-07,
+      "loss": 0.5009,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7024,
+      "grad_norm": 1.011412683899952,
+      "learning_rate": 2.9695390497770535e-07,
+      "loss": 0.6488,
+      "step": 1689
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.9133387908486027,
+      "learning_rate": 2.93800997204271e-07,
+      "loss": 0.5873,
+      "step": 1690
+    },
+    {
+      "epoch": 2.7056,
+      "grad_norm": 0.9545062457004707,
+      "learning_rate": 2.9066441016692594e-07,
+      "loss": 0.5496,
+      "step": 1691
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 1.0166317184941516,
+      "learning_rate": 2.8754415474311235e-07,
+      "loss": 0.7333,
+      "step": 1692
+    },
+    {
+      "epoch": 2.7088,
+      "grad_norm": 0.7225584678692126,
+      "learning_rate": 2.844402417536374e-07,
+      "loss": 0.3224,
+      "step": 1693
+    },
+    {
+      "epoch": 2.7104,
+      "grad_norm": 0.8868796779774393,
+      "learning_rate": 2.8135268196263055e-07,
+      "loss": 0.5455,
+      "step": 1694
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 0.7639113189038628,
+      "learning_rate": 2.782814860775124e-07,
+      "loss": 0.3971,
+      "step": 1695
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 1.0463004434919674,
+      "learning_rate": 2.752266647489549e-07,
+      "loss": 0.6093,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7152,
+      "grad_norm": 0.9661113131018776,
+      "learning_rate": 2.7218822857084217e-07,
+      "loss": 0.4983,
+      "step": 1697
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.9291352513206746,
+      "learning_rate": 2.691661880802382e-07,
+      "loss": 0.5323,
+      "step": 1698
+    },
+    {
+      "epoch": 2.7184,
+      "grad_norm": 0.7842018041059153,
+      "learning_rate": 2.661605537573453e-07,
+      "loss": 0.3933,
+      "step": 1699
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.6695511009471947,
+      "learning_rate": 2.631713360254734e-07,
+      "loss": 0.357,
+      "step": 1700
+    },
+    {
+      "epoch": 2.7216,
+      "grad_norm": 0.9752728154251148,
+      "learning_rate": 2.6019854525099977e-07,
+      "loss": 0.609,
+      "step": 1701
+    },
+    {
+      "epoch": 2.7232,
+      "grad_norm": 0.9003259467346113,
+      "learning_rate": 2.572421917433332e-07,
+      "loss": 0.4164,
+      "step": 1702
+    },
+    {
+      "epoch": 2.7248,
+      "grad_norm": 0.6820123456128315,
+      "learning_rate": 2.5430228575488156e-07,
+      "loss": 0.3883,
+      "step": 1703
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.8142029207278232,
+      "learning_rate": 2.513788374810111e-07,
+      "loss": 0.4403,
+      "step": 1704
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 1.0114785902582635,
+      "learning_rate": 2.4847185706001643e-07,
+      "loss": 0.5507,
+      "step": 1705
+    },
+    {
+      "epoch": 2.7296,
+      "grad_norm": 1.031565197570965,
+      "learning_rate": 2.455813545730812e-07,
+      "loss": 0.3876,
+      "step": 1706
+    },
+    {
+      "epoch": 2.7312,
+      "grad_norm": 0.9973880785202093,
+      "learning_rate": 2.4270734004424643e-07,
+      "loss": 0.5462,
+      "step": 1707
+    },
+    {
+      "epoch": 2.7328,
+      "grad_norm": 1.0928308631380754,
+      "learning_rate": 2.39849823440374e-07,
+      "loss": 0.404,
+      "step": 1708
+    },
+    {
+      "epoch": 2.7344,
+      "grad_norm": 0.867317346855674,
+      "learning_rate": 2.3700881467111025e-07,
+      "loss": 0.4486,
+      "step": 1709
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 1.0334464401782197,
+      "learning_rate": 2.3418432358885633e-07,
+      "loss": 0.5629,
+      "step": 1710
+    },
+    {
+      "epoch": 2.7376,
+      "grad_norm": 0.7909830652600024,
+      "learning_rate": 2.3137635998872808e-07,
+      "loss": 0.4398,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 1.2832617530267387,
+      "learning_rate": 2.285849336085294e-07,
+      "loss": 0.6437,
+      "step": 1712
+    },
+    {
+      "epoch": 2.7408,
+      "grad_norm": 0.8836239713948657,
+      "learning_rate": 2.258100541287117e-07,
+      "loss": 0.6076,
+      "step": 1713
+    },
+    {
+      "epoch": 2.7424,
+      "grad_norm": 1.0857587768731678,
+      "learning_rate": 2.2305173117234236e-07,
+      "loss": 0.5816,
+      "step": 1714
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 0.8472083599450355,
+      "learning_rate": 2.2030997430507462e-07,
+      "loss": 0.4962,
+      "step": 1715
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.8434014834262357,
+      "learning_rate": 2.1758479303510937e-07,
+      "loss": 0.5518,
+      "step": 1716
+    },
+    {
+      "epoch": 2.7472,
+      "grad_norm": 1.1849408576166802,
+      "learning_rate": 2.148761968131663e-07,
+      "loss": 0.5953,
+      "step": 1717
+    },
+    {
+      "epoch": 2.7488,
+      "grad_norm": 0.8553362198678027,
+      "learning_rate": 2.121841950324488e-07,
+      "loss": 0.5281,
+      "step": 1718
+    },
+    {
+      "epoch": 2.7504,
+      "grad_norm": 0.8558389777192571,
+      "learning_rate": 2.0950879702861082e-07,
+      "loss": 0.5287,
+      "step": 1719
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.9765310417692751,
+      "learning_rate": 2.0685001207972843e-07,
+      "loss": 0.5537,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7536,
+      "grad_norm": 0.9436193099371826,
+      "learning_rate": 2.042078494062616e-07,
+      "loss": 0.4218,
+      "step": 1721
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.8355901313855413,
+      "learning_rate": 2.0158231817102858e-07,
+      "loss": 0.5243,
+      "step": 1722
+    },
+    {
+      "epoch": 2.7568,
+      "grad_norm": 0.8864902252071198,
+      "learning_rate": 1.9897342747916938e-07,
+      "loss": 0.4518,
+      "step": 1723
+    },
+    {
+      "epoch": 2.7584,
+      "grad_norm": 0.8324001161614192,
+      "learning_rate": 1.9638118637811564e-07,
+      "loss": 0.5128,
+      "step": 1724
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.9284958368739817,
+      "learning_rate": 1.9380560385756088e-07,
+      "loss": 0.4961,
+      "step": 1725
+    },
+    {
+      "epoch": 2.7616,
+      "grad_norm": 0.9950902011711322,
+      "learning_rate": 1.9124668884942632e-07,
+      "loss": 0.6141,
+      "step": 1726
+    },
+    {
+      "epoch": 2.7632,
+      "grad_norm": 0.8401272664041367,
+      "learning_rate": 1.8870445022783234e-07,
+      "loss": 0.3772,
+      "step": 1727
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.952108716456919,
+      "learning_rate": 1.861788968090683e-07,
+      "loss": 0.591,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7664,
+      "grad_norm": 1.0321321969178023,
+      "learning_rate": 1.8367003735155764e-07,
+      "loss": 0.5918,
+      "step": 1729
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 0.791821254259906,
+      "learning_rate": 1.8117788055583286e-07,
+      "loss": 0.4422,
+      "step": 1730
+    },
+    {
+      "epoch": 2.7696,
+      "grad_norm": 0.8173771066344215,
+      "learning_rate": 1.7870243506450113e-07,
+      "loss": 0.4135,
+      "step": 1731
+    },
+    {
+      "epoch": 2.7712,
+      "grad_norm": 0.9186841190744106,
+      "learning_rate": 1.762437094622177e-07,
+      "loss": 0.4959,
+      "step": 1732
+    },
+    {
+      "epoch": 2.7728,
+      "grad_norm": 1.0275537017373018,
+      "learning_rate": 1.738017122756541e-07,
+      "loss": 0.6584,
+      "step": 1733
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 1.0430161165003917,
+      "learning_rate": 1.713764519734673e-07,
+      "loss": 0.642,
+      "step": 1734
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 0.9567526786143894,
+      "learning_rate": 1.68967936966275e-07,
+      "loss": 0.5208,
+      "step": 1735
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.812487127383462,
+      "learning_rate": 1.6657617560662088e-07,
+      "loss": 0.3675,
+      "step": 1736
+    },
+    {
+      "epoch": 2.7792,
+      "grad_norm": 0.9373211060381809,
+      "learning_rate": 1.6420117618895003e-07,
+      "loss": 0.581,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7808,
+      "grad_norm": 0.9841315757972408,
+      "learning_rate": 1.6184294694957747e-07,
+      "loss": 0.6086,
+      "step": 1738
+    },
+    {
+      "epoch": 2.7824,
+      "grad_norm": 0.8791421198943066,
+      "learning_rate": 1.5950149606666077e-07,
+      "loss": 0.4569,
+      "step": 1739
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.9642648167692756,
+      "learning_rate": 1.5717683166017184e-07,
+      "loss": 0.5504,
+      "step": 1740
+    },
+    {
+      "epoch": 2.7856,
+      "grad_norm": 0.9119074242238372,
+      "learning_rate": 1.5486896179186693e-07,
+      "loss": 0.5595,
+      "step": 1741
+    },
+    {
+      "epoch": 2.7872,
+      "grad_norm": 0.8971541475887924,
+      "learning_rate": 1.5257789446526172e-07,
+      "loss": 0.5563,
+      "step": 1742
+    },
+    {
+      "epoch": 2.7888,
+      "grad_norm": 0.8042412356157594,
+      "learning_rate": 1.5030363762560228e-07,
+      "loss": 0.4276,
+      "step": 1743
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 1.0800812712251564,
+      "learning_rate": 1.480461991598353e-07,
+      "loss": 0.7225,
+      "step": 1744
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 1.0672625382493286,
+      "learning_rate": 1.458055868965841e-07,
+      "loss": 0.5413,
+      "step": 1745
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.8877356286665732,
+      "learning_rate": 1.4358180860611913e-07,
+      "loss": 0.4827,
+      "step": 1746
+    },
+    {
+      "epoch": 2.7952,
+      "grad_norm": 0.8410448434984354,
+      "learning_rate": 1.4137487200033383e-07,
+      "loss": 0.4581,
+      "step": 1747
+    },
+    {
+      "epoch": 2.7968,
+      "grad_norm": 1.0506355864583343,
+      "learning_rate": 1.3918478473271325e-07,
+      "loss": 0.6866,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7984,
+      "grad_norm": 0.8777940704455622,
+      "learning_rate": 1.3701155439831249e-07,
+      "loss": 0.517,
+      "step": 1749
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.051779444323007,
+      "learning_rate": 1.3485518853372625e-07,
+      "loss": 0.6553,
+      "step": 1750
+    },
+    {
+      "epoch": 2.8016,
+      "grad_norm": 0.9332180897974,
+      "learning_rate": 1.3271569461706547e-07,
+      "loss": 0.6018,
+      "step": 1751
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.9622031801842292,
+      "learning_rate": 1.305930800679317e-07,
+      "loss": 0.5358,
+      "step": 1752
+    },
+    {
+      "epoch": 2.8048,
+      "grad_norm": 0.9439169279836743,
+      "learning_rate": 1.2848735224738729e-07,
+      "loss": 0.6348,
+      "step": 1753
+    },
+    {
+      "epoch": 2.8064,
+      "grad_norm": 0.6141565699315273,
+      "learning_rate": 1.2639851845793583e-07,
+      "loss": 0.2859,
+      "step": 1754
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 0.9887019457101981,
+      "learning_rate": 1.2432658594349113e-07,
+      "loss": 0.503,
+      "step": 1755
+    },
+    {
+      "epoch": 2.8096,
+      "grad_norm": 0.755573233731766,
+      "learning_rate": 1.2227156188935552e-07,
+      "loss": 0.3755,
+      "step": 1756
+    },
+    {
+      "epoch": 2.8112,
+      "grad_norm": 0.8068516187741288,
+      "learning_rate": 1.202334534221955e-07,
+      "loss": 0.4035,
+      "step": 1757
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 1.0541894091692041,
+      "learning_rate": 1.1821226761001391e-07,
+      "loss": 0.5488,
+      "step": 1758
+    },
+    {
+      "epoch": 2.8144,
+      "grad_norm": 0.8215663663964111,
+      "learning_rate": 1.1620801146212723e-07,
+      "loss": 0.483,
+      "step": 1759
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 1.0599170394946014,
+      "learning_rate": 1.1422069192914221e-07,
+      "loss": 0.6629,
+      "step": 1760
+    },
+    {
+      "epoch": 2.8176,
+      "grad_norm": 1.0241069984777655,
+      "learning_rate": 1.1225031590292923e-07,
+      "loss": 0.6546,
+      "step": 1761
+    },
+    {
+      "epoch": 2.8192,
+      "grad_norm": 0.9567436147461478,
+      "learning_rate": 1.1029689021660183e-07,
+      "loss": 0.4679,
+      "step": 1762
+    },
+    {
+      "epoch": 2.8208,
+      "grad_norm": 1.0237155632844601,
+      "learning_rate": 1.0836042164448945e-07,
+      "loss": 0.5878,
+      "step": 1763
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.9145183098750288,
+      "learning_rate": 1.0644091690211633e-07,
+      "loss": 0.4294,
+      "step": 1764
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 1.025294366320079,
+      "learning_rate": 1.0453838264617711e-07,
+      "loss": 0.5874,
+      "step": 1765
+    },
+    {
+      "epoch": 2.8256,
+      "grad_norm": 0.8145556035089478,
+      "learning_rate": 1.0265282547451405e-07,
+      "loss": 0.5206,
+      "step": 1766
+    },
+    {
+      "epoch": 2.8272,
+      "grad_norm": 1.033388081285144,
+      "learning_rate": 1.0078425192609487e-07,
+      "loss": 0.7222,
+      "step": 1767
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.9383179711094793,
+      "learning_rate": 9.893266848098826e-08,
+      "loss": 0.6028,
+      "step": 1768
+    },
+    {
+      "epoch": 2.8304,
+      "grad_norm": 0.9219171426373641,
+      "learning_rate": 9.709808156034394e-08,
+      "loss": 0.4348,
+      "step": 1769
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.9775727388399273,
+      "learning_rate": 9.528049752636714e-08,
+      "loss": 0.5827,
+      "step": 1770
+    },
+    {
+      "epoch": 2.8336,
+      "grad_norm": 0.915416296587186,
+      "learning_rate": 9.347992268230022e-08,
+      "loss": 0.5401,
+      "step": 1771
+    },
+    {
+      "epoch": 2.8352,
+      "grad_norm": 1.0055777347226402,
+      "learning_rate": 9.169636327239883e-08,
+      "loss": 0.3765,
+      "step": 1772
+    },
+    {
+      "epoch": 2.8368,
+      "grad_norm": 0.8967909547614343,
+      "learning_rate": 8.992982548190809e-08,
+      "loss": 0.4848,
+      "step": 1773
+    },
+    {
+      "epoch": 2.8384,
+      "grad_norm": 0.9335655966624374,
+      "learning_rate": 8.818031543704641e-08,
+      "loss": 0.452,
+      "step": 1774
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.0392399356550421,
+      "learning_rate": 8.644783920498001e-08,
+      "loss": 0.5388,
+      "step": 1775
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.9289874520348161,
+      "learning_rate": 8.473240279380235e-08,
+      "loss": 0.5201,
+      "step": 1776
+    },
+    {
+      "epoch": 2.8432,
+      "grad_norm": 0.8453655211817417,
+      "learning_rate": 8.303401215251583e-08,
+      "loss": 0.3835,
+      "step": 1777
+    },
+    {
+      "epoch": 2.8448,
+      "grad_norm": 0.9480733448652274,
+      "learning_rate": 8.135267317100792e-08,
+      "loss": 0.4522,
+      "step": 1778
+    },
+    {
+      "epoch": 2.8464,
+      "grad_norm": 0.922483436866371,
+      "learning_rate": 7.968839168003395e-08,
+      "loss": 0.5468,
+      "step": 1779
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 0.8044570303909396,
+      "learning_rate": 7.804117345119266e-08,
+      "loss": 0.4441,
+      "step": 1780
+    },
+    {
+      "epoch": 2.8496,
+      "grad_norm": 0.9769896374917257,
+      "learning_rate": 7.64110241969107e-08,
+      "loss": 0.537,
+      "step": 1781
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 1.0307479199926861,
+      "learning_rate": 7.479794957042041e-08,
+      "loss": 0.6612,
+      "step": 1782
+    },
+    {
+      "epoch": 2.8528000000000002,
+      "grad_norm": 0.808217262963437,
+      "learning_rate": 7.320195516574036e-08,
+      "loss": 0.5036,
+      "step": 1783
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 1.044456028781398,
+      "learning_rate": 7.16230465176565e-08,
+      "loss": 0.5555,
+      "step": 1784
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 1.0239175798105442,
+      "learning_rate": 7.00612291017022e-08,
+      "loss": 0.4934,
+      "step": 1785
+    },
+    {
+      "epoch": 2.8576,
+      "grad_norm": 0.9293490059098942,
+      "learning_rate": 6.851650833414103e-08,
+      "loss": 0.6342,
+      "step": 1786
+    },
+    {
+      "epoch": 2.8592,
+      "grad_norm": 0.902262463643105,
+      "learning_rate": 6.698888957194505e-08,
+      "loss": 0.4887,
+      "step": 1787
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.9041811208749048,
+      "learning_rate": 6.547837811277824e-08,
+      "loss": 0.4734,
+      "step": 1788
+    },
+    {
+      "epoch": 2.8624,
+      "grad_norm": 1.0173774405665454,
+      "learning_rate": 6.39849791949787e-08,
+      "loss": 0.5473,
+      "step": 1789
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.9337482598178692,
+      "learning_rate": 6.250869799753866e-08,
+      "loss": 0.639,
+      "step": 1790
+    },
+    {
+      "epoch": 2.8656,
+      "grad_norm": 1.0814103881812365,
+      "learning_rate": 6.104953964008897e-08,
+      "loss": 0.5582,
+      "step": 1791
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 1.2094547894118997,
+      "learning_rate": 5.960750918287627e-08,
+      "loss": 0.3622,
+      "step": 1792
+    },
+    {
+      "epoch": 2.8688000000000002,
+      "grad_norm": 0.6464899543796374,
+      "learning_rate": 5.818261162675309e-08,
+      "loss": 0.2707,
+      "step": 1793
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.9514938041621019,
+      "learning_rate": 5.677485191315391e-08,
+      "loss": 0.5766,
+      "step": 1794
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 1.0670141743034245,
+      "learning_rate": 5.538423492408129e-08,
+      "loss": 0.4951,
+      "step": 1795
+    },
+    {
+      "epoch": 2.8736,
+      "grad_norm": 0.9576950385760803,
+      "learning_rate": 5.401076548208761e-08,
+      "loss": 0.6705,
+      "step": 1796
+    },
+    {
+      "epoch": 2.8752,
+      "grad_norm": 0.9553166311093505,
+      "learning_rate": 5.265444835025946e-08,
+      "loss": 0.4376,
+      "step": 1797
+    },
+    {
+      "epoch": 2.8768000000000002,
+      "grad_norm": 0.9085640110287027,
+      "learning_rate": 5.1315288232201e-08,
+      "loss": 0.6372,
+      "step": 1798
+    },
+    {
+      "epoch": 2.8784,
+      "grad_norm": 0.8032912680905456,
+      "learning_rate": 4.9993289772015116e-08,
+      "loss": 0.4572,
+      "step": 1799
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.7552822476251557,
+      "learning_rate": 4.8688457554291746e-08,
+      "loss": 0.374,
+      "step": 1800
+    },
+    {
+      "epoch": 2.8816,
+      "grad_norm": 0.86223196440621,
+      "learning_rate": 4.7400796104088434e-08,
+      "loss": 0.4994,
+      "step": 1801
+    },
+    {
+      "epoch": 2.8832,
+      "grad_norm": 0.8675334598925721,
+      "learning_rate": 4.613030988691536e-08,
+      "loss": 0.4502,
+      "step": 1802
+    },
+    {
+      "epoch": 2.8848000000000003,
+      "grad_norm": 0.8508840629804357,
+      "learning_rate": 4.4877003308722575e-08,
+      "loss": 0.3946,
+      "step": 1803
+    },
+    {
+      "epoch": 2.8864,
+      "grad_norm": 0.824632873492589,
+      "learning_rate": 4.364088071587891e-08,
+      "loss": 0.4637,
+      "step": 1804
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 0.9845171821161611,
+      "learning_rate": 4.2421946395164174e-08,
+      "loss": 0.6362,
+      "step": 1805
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.9939173511971835,
+      "learning_rate": 4.1220204573747534e-08,
+      "loss": 0.6751,
+      "step": 1806
+    },
+    {
+      "epoch": 2.8912,
+      "grad_norm": 1.0491033709853865,
+      "learning_rate": 4.0035659419178086e-08,
+      "loss": 0.6523,
+      "step": 1807
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.887885585545425,
+      "learning_rate": 3.88683150393665e-08,
+      "loss": 0.451,
+      "step": 1808
+    },
+    {
+      "epoch": 2.8944,
+      "grad_norm": 0.9186365473940187,
+      "learning_rate": 3.771817548257395e-08,
+      "loss": 0.5534,
+      "step": 1809
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 0.8916838733612182,
+      "learning_rate": 3.658524473739544e-08,
+      "loss": 0.5114,
+      "step": 1810
+    },
+    {
+      "epoch": 2.8975999999999997,
+      "grad_norm": 0.9291647540672928,
+      "learning_rate": 3.546952673274817e-08,
+      "loss": 0.5271,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.9400922485383789,
+      "learning_rate": 3.437102533785541e-08,
+      "loss": 0.6075,
+      "step": 1812
+    },
+    {
+      "epoch": 2.9008000000000003,
+      "grad_norm": 0.8671931634607897,
+      "learning_rate": 3.328974436223709e-08,
+      "loss": 0.5047,
+      "step": 1813
+    },
+    {
+      "epoch": 2.9024,
+      "grad_norm": 0.9670018877245997,
+      "learning_rate": 3.2225687555690886e-08,
+      "loss": 0.7172,
+      "step": 1814
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 0.8860508394198572,
+      "learning_rate": 3.117885860828396e-08,
+      "loss": 0.5361,
+      "step": 1815
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.8849856446629526,
+      "learning_rate": 3.014926115034012e-08,
+      "loss": 0.5887,
+      "step": 1816
+    },
+    {
+      "epoch": 2.9072,
+      "grad_norm": 0.8878094619082265,
+      "learning_rate": 2.9136898752422648e-08,
+      "loss": 0.4742,
+      "step": 1817
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 1.0116183127223932,
+      "learning_rate": 2.8141774925327103e-08,
+      "loss": 0.6907,
+      "step": 1818
+    },
+    {
+      "epoch": 2.9104,
+      "grad_norm": 0.9214430805037775,
+      "learning_rate": 2.7163893120066288e-08,
+      "loss": 0.5466,
+      "step": 1819
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 0.8421919544242136,
+      "learning_rate": 2.6203256727859172e-08,
+      "loss": 0.4023,
+      "step": 1820
+    },
+    {
+      "epoch": 2.9135999999999997,
+      "grad_norm": 0.8216781699994113,
+      "learning_rate": 2.5259869080118127e-08,
+      "loss": 0.4008,
+      "step": 1821
+    },
+    {
+      "epoch": 2.9152,
+      "grad_norm": 0.9800442506370086,
+      "learning_rate": 2.4333733448440033e-08,
+      "loss": 0.577,
+      "step": 1822
+    },
+    {
+      "epoch": 2.9168,
+      "grad_norm": 1.223023270543174,
+      "learning_rate": 2.34248530445913e-08,
+      "loss": 0.6786,
+      "step": 1823
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 1.0292904741803353,
+      "learning_rate": 2.2533231020499536e-08,
+      "loss": 0.6278,
+      "step": 1824
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.9575536459093767,
+      "learning_rate": 2.1658870468241332e-08,
+      "loss": 0.5548,
+      "step": 1825
+    },
+    {
+      "epoch": 2.9215999999999998,
+      "grad_norm": 0.9112409992594211,
+      "learning_rate": 2.0801774420031172e-08,
+      "loss": 0.4765,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9232,
+      "grad_norm": 0.9272134004581282,
+      "learning_rate": 1.9961945848213092e-08,
+      "loss": 0.3256,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9248,
+      "grad_norm": 0.7604191272311528,
+      "learning_rate": 1.9139387665247922e-08,
+      "loss": 0.4347,
+      "step": 1828
+    },
+    {
+      "epoch": 2.9264,
+      "grad_norm": 0.9076059883586481,
+      "learning_rate": 1.8334102723703286e-08,
+      "loss": 0.4715,
+      "step": 1829
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 1.0102556282706276,
+      "learning_rate": 1.754609381624639e-08,
+      "loss": 0.6792,
+      "step": 1830
+    },
+    {
+      "epoch": 2.9295999999999998,
+      "grad_norm": 0.9182388853498075,
+      "learning_rate": 1.677536367563126e-08,
+      "loss": 0.4526,
+      "step": 1831
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 1.0821328510966655,
+      "learning_rate": 1.6021914974690413e-08,
+      "loss": 0.5027,
+      "step": 1832
+    },
+    {
+      "epoch": 2.9328,
+      "grad_norm": 1.0481489569364173,
+      "learning_rate": 1.5285750326325953e-08,
+      "loss": 0.7295,
+      "step": 1833
+    },
+    {
+      "epoch": 2.9344,
+      "grad_norm": 1.0058485179008962,
+      "learning_rate": 1.4566872283500733e-08,
+      "loss": 0.636,
+      "step": 1834
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 1.0412473663197401,
+      "learning_rate": 1.3865283339228319e-08,
+      "loss": 0.5516,
+      "step": 1835
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.7609252138523108,
+      "learning_rate": 1.3180985926564693e-08,
+      "loss": 0.4838,
+      "step": 1836
+    },
+    {
+      "epoch": 2.9392,
+      "grad_norm": 1.1032529717834951,
+      "learning_rate": 1.2513982418601024e-08,
+      "loss": 0.5476,
+      "step": 1837
+    },
+    {
+      "epoch": 2.9408,
+      "grad_norm": 1.0831922971527597,
+      "learning_rate": 1.1864275128454783e-08,
+      "loss": 0.4526,
+      "step": 1838
+    },
+    {
+      "epoch": 2.9424,
+      "grad_norm": 0.7400506166918026,
+      "learning_rate": 1.1231866309259764e-08,
+      "loss": 0.3346,
+      "step": 1839
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.9640678592204933,
+      "learning_rate": 1.0616758154161633e-08,
+      "loss": 0.559,
+      "step": 1840
+    },
+    {
+      "epoch": 2.9455999999999998,
+      "grad_norm": 0.8982123231849045,
+      "learning_rate": 1.0018952796307934e-08,
+      "loss": 0.5302,
+      "step": 1841
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.941060865328355,
+      "learning_rate": 9.438452308841995e-09,
+      "loss": 0.5808,
+      "step": 1842
+    },
+    {
+      "epoch": 2.9488,
+      "grad_norm": 0.8803813607496823,
+      "learning_rate": 8.87525870489514e-09,
+      "loss": 0.5254,
+      "step": 1843
+    },
+    {
+      "epoch": 2.9504,
+      "grad_norm": 1.1169655409683366,
+      "learning_rate": 8.329373937578378e-09,
+      "loss": 0.6741,
+      "step": 1844
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 0.9707685846711582,
+      "learning_rate": 7.800799899979061e-09,
+      "loss": 0.5368,
+      "step": 1845
+    },
+    {
+      "epoch": 2.9536,
+      "grad_norm": 1.0657318722904765,
+      "learning_rate": 7.289538425150899e-09,
+      "loss": 0.5905,
+      "step": 1846
+    },
+    {
+      "epoch": 2.9552,
+      "grad_norm": 0.9452865407104821,
+      "learning_rate": 6.7955912861095155e-09,
+      "loss": 0.5977,
+      "step": 1847
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.9118312275808697,
+      "learning_rate": 6.31896019582523e-09,
+      "loss": 0.571,
+      "step": 1848
+    },
+    {
+      "epoch": 2.9584,
+      "grad_norm": 1.0299071317638093,
+      "learning_rate": 5.8596468072180665e-09,
+      "loss": 0.6393,
+      "step": 1849
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7897766327775042,
+      "learning_rate": 5.417652713152199e-09,
+      "loss": 0.4784,
+      "step": 1850
+    },
+    {
+      "epoch": 2.9616,
+      "grad_norm": 0.7981199203915393,
+      "learning_rate": 4.992979446428736e-09,
+      "loss": 0.4134,
+      "step": 1851
+    },
+    {
+      "epoch": 2.9632,
+      "grad_norm": 1.187210869908459,
+      "learning_rate": 4.585628479781279e-09,
+      "loss": 0.4586,
+      "step": 1852
+    },
+    {
+      "epoch": 2.9648,
+      "grad_norm": 0.6960407653490429,
+      "learning_rate": 4.195601225872592e-09,
+      "loss": 0.3188,
+      "step": 1853
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.9992176871297235,
+      "learning_rate": 3.822899037286276e-09,
+      "loss": 0.4929,
+      "step": 1854
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 1.0207306620965482,
+      "learning_rate": 3.4675232065256583e-09,
+      "loss": 0.5049,
+      "step": 1855
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.9015786606691766,
+      "learning_rate": 3.129474966006574e-09,
+      "loss": 0.4999,
+      "step": 1856
+    },
+    {
+      "epoch": 2.9712,
+      "grad_norm": 0.9706573615393184,
+      "learning_rate": 2.808755488054038e-09,
+      "loss": 0.5992,
+      "step": 1857
+    },
+    {
+      "epoch": 2.9728,
+      "grad_norm": 0.7234137116143059,
+      "learning_rate": 2.5053658848989137e-09,
+      "loss": 0.3747,
+      "step": 1858
+    },
+    {
+      "epoch": 2.9744,
+      "grad_norm": 1.1051929123420487,
+      "learning_rate": 2.219307208672361e-09,
+      "loss": 0.6691,
+      "step": 1859
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.8735024834970568,
+      "learning_rate": 1.9505804514047266e-09,
+      "loss": 0.4822,
+      "step": 1860
+    },
+    {
+      "epoch": 2.9776,
+      "grad_norm": 1.0768083864366322,
+      "learning_rate": 1.6991865450188827e-09,
+      "loss": 0.6249,
+      "step": 1861
+    },
+    {
+      "epoch": 2.9792,
+      "grad_norm": 1.0894270482480275,
+      "learning_rate": 1.465126361330227e-09,
+      "loss": 0.6996,
+      "step": 1862
+    },
+    {
+      "epoch": 2.9808,
+      "grad_norm": 0.9655612363772357,
+      "learning_rate": 1.2484007120411312e-09,
+      "loss": 0.5698,
+      "step": 1863
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.9616266425429173,
+      "learning_rate": 1.0490103487392766e-09,
+      "loss": 0.5962,
+      "step": 1864
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 1.040108530522099,
+      "learning_rate": 8.669559628954327e-10,
+      "loss": 0.3869,
+      "step": 1865
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.8586655065163483,
+      "learning_rate": 7.02238185860682e-10,
+      "loss": 0.522,
+      "step": 1866
+    },
+    {
+      "epoch": 2.9872,
+      "grad_norm": 0.8400689514126244,
+      "learning_rate": 5.54857588862534e-10,
+      "loss": 0.4692,
+      "step": 1867
+    },
+    {
+      "epoch": 2.9888,
+      "grad_norm": 1.0257772892218056,
+      "learning_rate": 4.2481468300603625e-10,
+      "loss": 0.6331,
+      "step": 1868
+    },
+    {
+      "epoch": 2.9904,
+      "grad_norm": 1.0477332250877671,
+      "learning_rate": 3.1210991927044244e-10,
+      "loss": 0.5845,
+      "step": 1869
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.8218386967421416,
+      "learning_rate": 2.167436885064378e-10,
+      "loss": 0.5359,
+      "step": 1870
+    },
+    {
+      "epoch": 2.9936,
+      "grad_norm": 0.9216931045691216,
+      "learning_rate": 1.387163214372489e-10,
+      "loss": 0.3568,
+      "step": 1871
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.7960561651454994,
+      "learning_rate": 7.80280886558682e-11,
+      "loss": 0.4585,
+      "step": 1872
+    },
+    {
+      "epoch": 2.9968,
+      "grad_norm": 0.854304909671456,
+      "learning_rate": 3.467920062394381e-11,
+      "loss": 0.4056,
+      "step": 1873
+    },
+    {
+      "epoch": 2.9984,
+      "grad_norm": 0.7788636400445171,
+      "learning_rate": 8.669807672334606e-12,
+      "loss": 0.4405,
+      "step": 1874
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9117903366366832,
+      "learning_rate": 0.0,
+      "loss": 0.4754,
+      "step": 1875
+    },
+    {
+      "epoch": 3.0,
+      "step": 1875,
+      "total_flos": 1060946637111296.0,
+      "train_loss": 0.7549448100566865,
+      "train_runtime": 19284.8604,
+      "train_samples_per_second": 1.556,
+      "train_steps_per_second": 0.097
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1875,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1060946637111296.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}