diff --git "a/trainer_state.json" "b/trainer_state.json"
deleted file mode 100755--- "a/trainer_state.json"
+++ /dev/null
@@ -1,13560 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.998172403289674,
-  "eval_steps": 25.0,
-  "global_step": 1230,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "full_loss": 0.6115,
-      "grad_norm": 11.4375,
-      "learning_rate": 6.756756756756758e-07,
-      "long_answer_loss": 0.6115,
-      "loss": 0.6482,
-      "short_answer_loss": NaN,
-      "step": 1,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.0,
-      "full_loss": 0.6372,
-      "grad_norm": 11.3125,
-      "learning_rate": 1.3513513513513515e-06,
-      "long_answer_loss": 0.6372,
-      "loss": 0.6202,
-      "short_answer_loss": NaN,
-      "step": 2,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.0,
-      "full_loss": 0.7199,
-      "grad_norm": 10.75,
-      "learning_rate": 2.0270270270270273e-06,
-      "long_answer_loss": 0.7199,
-      "loss": 0.6418,
-      "short_answer_loss": NaN,
-      "step": 3,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.6846,
-      "grad_norm": 10.625,
-      "learning_rate": 2.702702702702703e-06,
-      "long_answer_loss": 0.6846,
-      "loss": 0.6412,
-      "short_answer_loss": NaN,
-      "step": 4,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.5437,
-      "grad_norm": 9.5625,
-      "learning_rate": 3.3783783783783788e-06,
-      "long_answer_loss": 0.5437,
-      "loss": 0.5699,
-      "short_answer_loss": NaN,
-      "step": 5,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.5362,
-      "grad_norm": 8.3125,
-      "learning_rate": 4.0540540540540545e-06,
-      "long_answer_loss": 0.5362,
-      "loss": 0.5227,
-      "short_answer_loss": NaN,
-      "step": 6,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.4596,
-      "grad_norm": 7.0625,
-      "learning_rate": 4.72972972972973e-06,
-      "long_answer_loss": 0.4596,
-      "loss": 0.489,
-      "short_answer_loss": NaN,
-      "step": 7,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.3601,
-      "grad_norm": 5.3125,
-      "learning_rate": 5.405405405405406e-06,
-      "long_answer_loss": 0.3601,
-      "loss": 0.3993,
-      "short_answer_loss": NaN,
-      "step": 8,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.361,
-      "grad_norm": 6.03125,
-      "learning_rate": 6.081081081081082e-06,
-      "long_answer_loss": 0.361,
-      "loss": 0.3465,
-      "short_answer_loss": NaN,
-      "step": 9,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.3461,
-      "grad_norm": 5.78125,
-      "learning_rate": 6.7567567567567575e-06,
-      "long_answer_loss": 0.3461,
-      "loss": 0.3391,
-      "short_answer_loss": NaN,
-      "step": 10,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.3227,
-      "grad_norm": 5.0625,
-      "learning_rate": 7.432432432432433e-06,
-      "long_answer_loss": 0.3227,
-      "loss": 0.3084,
-      "short_answer_loss": NaN,
-      "step": 11,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2806,
-      "grad_norm": 4.25,
-      "learning_rate": 8.108108108108109e-06,
-      "long_answer_loss": 0.2806,
-      "loss": 0.3001,
-      "short_answer_loss": NaN,
-      "step": 12,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.3417,
-      "grad_norm": 3.46875,
-      "learning_rate": 8.783783783783785e-06,
-      "long_answer_loss": 0.3417,
-      "loss": 0.3107,
-      "short_answer_loss": NaN,
-      "step": 13,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.3048,
-      "grad_norm": 3.0625,
-      "learning_rate": 9.45945945945946e-06,
-      "long_answer_loss": 0.3048,
-      "loss": 0.2865,
-      "short_answer_loss": NaN,
-      "step": 14,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2859,
-      "grad_norm": 2.953125,
-      "learning_rate": 1.0135135135135136e-05,
-      "long_answer_loss": 0.2859,
-      "loss": 0.2784,
-      "short_answer_loss": NaN,
-      "step": 15,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.3344,
-      "grad_norm": 3.140625,
-      "learning_rate": 1.0810810810810812e-05,
-      "long_answer_loss": 0.3344,
-      "loss": 0.2624,
-      "short_answer_loss": NaN,
-      "step": 16,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2341,
-      "grad_norm": 3.15625,
-      "learning_rate": 1.1486486486486488e-05,
-      "long_answer_loss": 0.2341,
-      "loss": 0.2604,
-      "short_answer_loss": NaN,
-      "step": 17,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2536,
-      "grad_norm": 2.796875,
-      "learning_rate": 1.2162162162162164e-05,
-      "long_answer_loss": 0.2536,
-      "loss": 0.2468,
-      "short_answer_loss": NaN,
-      "step": 18,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2628,
-      "grad_norm": 2.828125,
-      "learning_rate": 1.2837837837837838e-05,
-      "long_answer_loss": 0.2628,
-      "loss": 0.2661,
-      "short_answer_loss": NaN,
-      "step": 19,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2042,
-      "grad_norm": 2.875,
-      "learning_rate": 1.3513513513513515e-05,
-      "long_answer_loss": 0.2042,
-      "loss": 0.2447,
-      "short_answer_loss": NaN,
-      "step": 20,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2151,
-      "grad_norm": 2.40625,
-      "learning_rate": 1.4189189189189189e-05,
-      "long_answer_loss": 0.2151,
-      "loss": 0.233,
-      "short_answer_loss": NaN,
-      "step": 21,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2608,
-      "grad_norm": 2.734375,
-      "learning_rate": 1.4864864864864867e-05,
-      "long_answer_loss": 0.2608,
-      "loss": 0.2464,
-      "short_answer_loss": NaN,
-      "step": 22,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.3018,
-      "grad_norm": 2.109375,
-      "learning_rate": 1.554054054054054e-05,
-      "long_answer_loss": 0.3018,
-      "loss": 0.2438,
-      "short_answer_loss": NaN,
-      "step": 23,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2322,
-      "grad_norm": 2.59375,
-      "learning_rate": 1.6216216216216218e-05,
-      "long_answer_loss": 0.2322,
-      "loss": 0.2321,
-      "short_answer_loss": NaN,
-      "step": 24,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2401,
-      "grad_norm": 2.40625,
-      "learning_rate": 1.6891891891891892e-05,
-      "long_answer_loss": 0.2401,
-      "loss": 0.2233,
-      "short_answer_loss": NaN,
-      "step": 25,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2226,
-      "grad_norm": 2.609375,
-      "learning_rate": 1.756756756756757e-05,
-      "long_answer_loss": 0.2226,
-      "loss": 0.2298,
-      "short_answer_loss": NaN,
-      "step": 26,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.1849,
-      "grad_norm": 2.5,
-      "learning_rate": 1.8243243243243244e-05,
-      "long_answer_loss": 0.1849,
-      "loss": 0.2131,
-      "short_answer_loss": NaN,
-      "step": 27,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2093,
-      "grad_norm": 2.421875,
-      "learning_rate": 1.891891891891892e-05,
-      "long_answer_loss": 0.2093,
-      "loss": 0.2247,
-      "short_answer_loss": NaN,
-      "step": 28,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2368,
-      "grad_norm": 2.59375,
-      "learning_rate": 1.9594594594594595e-05,
-      "long_answer_loss": 0.2368,
-      "loss": 0.2149,
-      "short_answer_loss": NaN,
-      "step": 29,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2029,
-      "grad_norm": 2.609375,
-      "learning_rate": 2.0270270270270273e-05,
-      "long_answer_loss": 0.2029,
-      "loss": 0.2181,
-      "short_answer_loss": NaN,
-      "step": 30,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2043,
-      "grad_norm": 2.484375,
-      "learning_rate": 2.0945945945945947e-05,
-      "long_answer_loss": 0.2043,
-      "loss": 0.2165,
-      "short_answer_loss": NaN,
-      "step": 31,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2517,
-      "grad_norm": 2.546875,
-      "learning_rate": 2.1621621621621624e-05,
-      "long_answer_loss": 0.2517,
-      "loss": 0.2321,
-      "short_answer_loss": NaN,
-      "step": 32,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2206,
-      "grad_norm": 2.484375,
-      "learning_rate": 2.2297297297297298e-05,
-      "long_answer_loss": 0.2206,
-      "loss": 0.2293,
-      "short_answer_loss": NaN,
-      "step": 33,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2023,
-      "grad_norm": 2.375,
-      "learning_rate": 2.2972972972972976e-05,
-      "long_answer_loss": 0.2023,
-      "loss": 0.2153,
-      "short_answer_loss": NaN,
-      "step": 34,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2392,
-      "grad_norm": 2.796875,
-      "learning_rate": 2.364864864864865e-05,
-      "long_answer_loss": 0.2392,
-      "loss": 0.2223,
-      "short_answer_loss": NaN,
-      "step": 35,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2382,
-      "grad_norm": 2.578125,
-      "learning_rate": 2.4324324324324327e-05,
-      "long_answer_loss": 0.2382,
-      "loss": 0.2365,
-      "short_answer_loss": NaN,
-      "step": 36,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2238,
-      "grad_norm": 2.375,
-      "learning_rate": 2.5e-05,
-      "long_answer_loss": 0.2238,
-      "loss": 0.2225,
-      "short_answer_loss": NaN,
-      "step": 37,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2399,
-      "grad_norm": 2.65625,
-      "learning_rate": 2.499995665903025e-05,
-      "long_answer_loss": 0.2399,
-      "loss": 0.2232,
-      "short_answer_loss": NaN,
-      "step": 38,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2196,
-      "grad_norm": 2.65625,
-      "learning_rate": 2.4999826636421537e-05,
-      "long_answer_loss": 0.2196,
-      "loss": 0.2163,
-      "short_answer_loss": NaN,
-      "step": 39,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.1898,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.4999609933075525e-05,
-      "long_answer_loss": 0.1898,
-      "loss": 0.2222,
-      "short_answer_loss": NaN,
-      "step": 40,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.2008,
-      "grad_norm": 2.6875,
-      "learning_rate": 2.4999306550494938e-05,
-      "long_answer_loss": 0.2008,
-      "loss": 0.2219,
-      "short_answer_loss": NaN,
-      "step": 41,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.2256,
-      "grad_norm": 2.921875,
-      "learning_rate": 2.4998916490783615e-05,
-      "long_answer_loss": 0.2256,
-      "loss": 0.2265,
-      "short_answer_loss": NaN,
-      "step": 42,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.2206,
-      "grad_norm": 2.640625,
-      "learning_rate": 2.499843975664644e-05,
-      "long_answer_loss": 0.2206,
-      "loss": 0.2157,
-      "short_answer_loss": NaN,
-      "step": 43,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.2298,
-      "grad_norm": 2.453125,
-      "learning_rate": 2.499787635138935e-05,
-      "long_answer_loss": 0.2298,
-      "loss": 0.2231,
-      "short_answer_loss": NaN,
-      "step": 44,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1787,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4997226278919313e-05,
-      "long_answer_loss": 0.1787,
-      "loss": 0.207,
-      "short_answer_loss": NaN,
-      "step": 45,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1969,
-      "grad_norm": 2.515625,
-      "learning_rate": 2.499648954374429e-05,
-      "long_answer_loss": 0.1969,
-      "loss": 0.2057,
-      "short_answer_loss": NaN,
-      "step": 46,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.2109,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.4995666150973213e-05,
-      "long_answer_loss": 0.2109,
-      "loss": 0.2189,
-      "short_answer_loss": NaN,
-      "step": 47,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.2601,
-      "grad_norm": 2.34375,
-      "learning_rate": 2.4994756106315946e-05,
-      "long_answer_loss": 0.2601,
-      "loss": 0.2189,
-      "short_answer_loss": NaN,
-      "step": 48,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1932,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.4993759416083243e-05,
-      "long_answer_loss": 0.1932,
-      "loss": 0.2043,
-      "short_answer_loss": NaN,
-      "step": 49,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.2208,
-      "grad_norm": 2.3125,
-      "learning_rate": 2.4992676087186707e-05,
-      "long_answer_loss": 0.2208,
-      "loss": 0.2145,
-      "short_answer_loss": NaN,
-      "step": 50,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.2448,
-      "grad_norm": 2.296875,
-      "learning_rate": 2.4991506127138743e-05,
-      "long_answer_loss": 0.2448,
-      "loss": 0.2071,
-      "short_answer_loss": NaN,
-      "step": 51,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1996,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.4990249544052495e-05,
-      "long_answer_loss": 0.1996,
-      "loss": 0.2221,
-      "short_answer_loss": NaN,
-      "step": 52,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2223,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.4988906346641824e-05,
-      "long_answer_loss": 0.2223,
-      "loss": 0.2183,
-      "short_answer_loss": NaN,
-      "step": 53,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2183,
-      "grad_norm": 2.421875,
-      "learning_rate": 2.4987476544221195e-05,
-      "long_answer_loss": 0.2183,
-      "loss": 0.2144,
-      "short_answer_loss": NaN,
-      "step": 54,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2327,
-      "grad_norm": 2.25,
-      "learning_rate": 2.4985960146705657e-05,
-      "long_answer_loss": 0.2327,
-      "loss": 0.2208,
-      "short_answer_loss": NaN,
-      "step": 55,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1777,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4984357164610743e-05,
-      "long_answer_loss": 0.1777,
-      "loss": 0.2029,
-      "short_answer_loss": NaN,
-      "step": 56,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2131,
-      "grad_norm": 2.359375,
-      "learning_rate": 2.4982667609052434e-05,
-      "long_answer_loss": 0.2131,
-      "loss": 0.2168,
-      "short_answer_loss": NaN,
-      "step": 57,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2355,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.4980891491747036e-05,
-      "long_answer_loss": 0.2355,
-      "loss": 0.228,
-      "short_answer_loss": NaN,
-      "step": 58,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2016,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.4979028825011137e-05,
-      "long_answer_loss": 0.2016,
-      "loss": 0.2088,
-      "short_answer_loss": NaN,
-      "step": 59,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.197,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.49770796217615e-05,
-      "long_answer_loss": 0.197,
-      "loss": 0.2219,
-      "short_answer_loss": NaN,
-      "step": 60,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2062,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4975043895514987e-05,
-      "long_answer_loss": 0.2062,
-      "loss": 0.2069,
-      "short_answer_loss": NaN,
-      "step": 61,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2231,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4972921660388448e-05,
-      "long_answer_loss": 0.2231,
-      "loss": 0.2046,
-      "short_answer_loss": NaN,
-      "step": 62,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2231,
-      "grad_norm": 2.25,
-      "learning_rate": 2.4970712931098644e-05,
-      "long_answer_loss": 0.2231,
-      "loss": 0.2077,
-      "short_answer_loss": NaN,
-      "step": 63,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2322,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.496841772296213e-05,
-      "long_answer_loss": 0.2322,
-      "loss": 0.2194,
-      "short_answer_loss": NaN,
-      "step": 64,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2073,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.4966036051895147e-05,
-      "long_answer_loss": 0.2073,
-      "loss": 0.2224,
-      "short_answer_loss": NaN,
-      "step": 65,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.1991,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4963567934413533e-05,
-      "long_answer_loss": 0.1991,
-      "loss": 0.2086,
-      "short_answer_loss": NaN,
-      "step": 66,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2302,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.4961013387632583e-05,
-      "long_answer_loss": 0.2302,
-      "loss": 0.2187,
-      "short_answer_loss": NaN,
-      "step": 67,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2221,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4958372429266934e-05,
-      "long_answer_loss": 0.2221,
-      "loss": 0.2174,
-      "short_answer_loss": NaN,
-      "step": 68,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2164,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.495564507763047e-05,
-      "long_answer_loss": 0.2164,
-      "loss": 0.2186,
-      "short_answer_loss": NaN,
-      "step": 69,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2571,
-      "grad_norm": 2.125,
-      "learning_rate": 2.495283135163615e-05,
-      "long_answer_loss": 0.2571,
-      "loss": 0.217,
-      "short_answer_loss": NaN,
-      "step": 70,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2136,
-      "grad_norm": 2.25,
-      "learning_rate": 2.494993127079592e-05,
-      "long_answer_loss": 0.2136,
-      "loss": 0.2178,
-      "short_answer_loss": NaN,
-      "step": 71,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.1974,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.494694485522055e-05,
-      "long_answer_loss": 0.1974,
-      "loss": 0.2117,
-      "short_answer_loss": NaN,
-      "step": 72,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2323,
-      "grad_norm": 2.125,
-      "learning_rate": 2.49438721256195e-05,
-      "long_answer_loss": 0.2323,
-      "loss": 0.2239,
-      "short_answer_loss": NaN,
-      "step": 73,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2445,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.4940713103300783e-05,
-      "long_answer_loss": 0.2445,
-      "loss": 0.2155,
-      "short_answer_loss": NaN,
-      "step": 74,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2139,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4937467810170818e-05,
-      "long_answer_loss": 0.2139,
-      "loss": 0.2217,
-      "short_answer_loss": NaN,
-      "step": 75,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.1922,
-      "grad_norm": 2.15625,
-      "learning_rate": 2.4934136268734265e-05,
-      "long_answer_loss": 0.1922,
-      "loss": 0.2067,
-      "short_answer_loss": NaN,
-      "step": 76,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.1909,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.493071850209388e-05,
-      "long_answer_loss": 0.1909,
-      "loss": 0.2108,
-      "short_answer_loss": NaN,
-      "step": 77,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2392,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4927214533950362e-05,
-      "long_answer_loss": 0.2392,
-      "loss": 0.2229,
-      "short_answer_loss": NaN,
-      "step": 78,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2388,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.4923624388602164e-05,
-      "long_answer_loss": 0.2388,
-      "loss": 0.2164,
-      "short_answer_loss": NaN,
-      "step": 79,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2281,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.491994809094535e-05,
-      "long_answer_loss": 0.2281,
-      "loss": 0.2121,
-      "short_answer_loss": NaN,
-      "step": 80,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2057,
-      "grad_norm": 2.21875,
-      "learning_rate": 2.4916185666473413e-05,
-      "long_answer_loss": 0.2057,
-      "loss": 0.2131,
-      "short_answer_loss": NaN,
-      "step": 81,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2526,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.4912337141277083e-05,
-      "long_answer_loss": 0.2526,
-      "loss": 0.2175,
-      "short_answer_loss": NaN,
-      "step": 82,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2265,
-      "grad_norm": 2.296875,
-      "learning_rate": 2.4908402542044178e-05,
-      "long_answer_loss": 0.2265,
-      "loss": 0.2176,
-      "short_answer_loss": NaN,
-      "step": 83,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.2284,
-      "grad_norm": 2.21875,
-      "learning_rate": 2.4904381896059393e-05,
-      "long_answer_loss": 0.2284,
-      "loss": 0.2216,
-      "short_answer_loss": NaN,
-      "step": 84,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.2304,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.490027523120412e-05,
-      "long_answer_loss": 0.2304,
-      "loss": 0.2233,
-      "short_answer_loss": NaN,
-      "step": 85,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.2291,
-      "grad_norm": 2.21875,
-      "learning_rate": 2.4896082575956242e-05,
-      "long_answer_loss": 0.2291,
-      "loss": 0.2229,
-      "short_answer_loss": NaN,
-      "step": 86,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1959,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4891803959389973e-05,
-      "long_answer_loss": 0.1959,
-      "loss": 0.2194,
-      "short_answer_loss": NaN,
-      "step": 87,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1872,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.4887439411175605e-05,
-      "long_answer_loss": 0.1872,
-      "loss": 0.2153,
-      "short_answer_loss": NaN,
-      "step": 88,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1849,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.488298896157935e-05,
-      "long_answer_loss": 0.1849,
-      "loss": 0.2242,
-      "short_answer_loss": NaN,
-      "step": 89,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1916,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.4878452641463083e-05,
-      "long_answer_loss": 0.1916,
-      "loss": 0.2008,
-      "short_answer_loss": NaN,
-      "step": 90,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2591,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.4873830482284173e-05,
-      "long_answer_loss": 0.2591,
-      "loss": 0.2244,
-      "short_answer_loss": NaN,
-      "step": 91,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2048,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.486912251609524e-05,
-      "long_answer_loss": 0.2048,
-      "loss": 0.2217,
-      "short_answer_loss": NaN,
-      "step": 92,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1927,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4864328775543927e-05,
-      "long_answer_loss": 0.1927,
-      "loss": 0.2027,
-      "short_answer_loss": NaN,
-      "step": 93,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2159,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.48594492938727e-05,
-      "long_answer_loss": 0.2159,
-      "loss": 0.2074,
-      "short_answer_loss": NaN,
-      "step": 94,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2118,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.485448410491859e-05,
-      "long_answer_loss": 0.2118,
-      "loss": 0.2206,
-      "short_answer_loss": NaN,
-      "step": 95,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2104,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4849433243112977e-05,
-      "long_answer_loss": 0.2104,
-      "loss": 0.2178,
-      "short_answer_loss": NaN,
-      "step": 96,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2231,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.4844296743481338e-05,
-      "long_answer_loss": 0.2231,
-      "loss": 0.2093,
-      "short_answer_loss": NaN,
-      "step": 97,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2079,
-      "grad_norm": 2.1875,
-      "learning_rate": 2.4839074641643012e-05,
-      "long_answer_loss": 0.2079,
-      "loss": 0.2246,
-      "short_answer_loss": NaN,
-      "step": 98,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1989,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.4833766973810953e-05,
-      "long_answer_loss": 0.1989,
-      "loss": 0.2094,
-      "short_answer_loss": NaN,
-      "step": 99,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2063,
-      "grad_norm": 2.0,
-      "learning_rate": 2.482837377679148e-05,
-      "long_answer_loss": 0.2063,
-      "loss": 0.2056,
-      "short_answer_loss": NaN,
-      "step": 100,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1934,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.482289508798401e-05,
-      "long_answer_loss": 0.1934,
-      "loss": 0.2093,
-      "short_answer_loss": NaN,
-      "step": 101,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.2258,
-      "grad_norm": 2.15625,
-      "learning_rate": 2.4817330945380817e-05,
-      "long_answer_loss": 0.2258,
-      "loss": 0.2083,
-      "short_answer_loss": NaN,
-      "step": 102,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1891,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.4811681387566755e-05,
-      "long_answer_loss": 0.1891,
-      "loss": 0.213,
-      "short_answer_loss": NaN,
-      "step": 103,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.202,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.4805946453718987e-05,
-      "long_answer_loss": 0.202,
-      "loss": 0.2108,
-      "short_answer_loss": NaN,
-      "step": 104,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.2068,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4800126183606735e-05,
-      "long_answer_loss": 0.2068,
-      "loss": 0.2098,
-      "short_answer_loss": NaN,
-      "step": 105,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.2159,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4794220617590985e-05,
-      "long_answer_loss": 0.2159,
-      "loss": 0.2147,
-      "short_answer_loss": NaN,
-      "step": 106,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.203,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.47882297966242e-05,
-      "long_answer_loss": 0.203,
-      "loss": 0.1986,
-      "short_answer_loss": NaN,
-      "step": 107,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.2326,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.478215376225007e-05,
-      "long_answer_loss": 0.2326,
-      "loss": 0.2143,
-      "short_answer_loss": NaN,
-      "step": 108,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.2194,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4775992556603188e-05,
-      "long_answer_loss": 0.2194,
-      "loss": 0.2204,
-      "short_answer_loss": NaN,
-      "step": 109,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1908,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.476974622240877e-05,
-      "long_answer_loss": 0.1908,
-      "loss": 0.2131,
-      "short_answer_loss": NaN,
-      "step": 110,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.2184,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.4763414802982364e-05,
-      "long_answer_loss": 0.2184,
-      "loss": 0.2103,
-      "short_answer_loss": NaN,
-      "step": 111,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1973,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.4756998342229552e-05,
-      "long_answer_loss": 0.1973,
-      "loss": 0.1993,
-      "short_answer_loss": NaN,
-      "step": 112,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.203,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.4750496884645637e-05,
-      "long_answer_loss": 0.203,
-      "loss": 0.2125,
-      "short_answer_loss": NaN,
-      "step": 113,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.2308,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.474391047531533e-05,
-      "long_answer_loss": 0.2308,
-      "loss": 0.2103,
-      "short_answer_loss": NaN,
-      "step": 114,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.176,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4737239159912452e-05,
-      "long_answer_loss": 0.176,
-      "loss": 0.2066,
-      "short_answer_loss": NaN,
-      "step": 115,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.2447,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4730482984699603e-05,
-      "long_answer_loss": 0.2447,
-      "loss": 0.213,
-      "short_answer_loss": NaN,
-      "step": 116,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1966,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4723641996527863e-05,
-      "long_answer_loss": 0.1966,
-      "loss": 0.2085,
-      "short_answer_loss": NaN,
-      "step": 117,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.2139,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.4716716242836432e-05,
-      "long_answer_loss": 0.2139,
-      "loss": 0.2043,
-      "short_answer_loss": NaN,
-      "step": 118,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.2305,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4709705771652336e-05,
-      "long_answer_loss": 0.2305,
-      "loss": 0.2182,
-      "short_answer_loss": NaN,
-      "step": 119,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1965,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.4702610631590073e-05,
-      "long_answer_loss": 0.1965,
-      "loss": 0.2,
-      "short_answer_loss": NaN,
-      "step": 120,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.234,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.4695430871851283e-05,
-      "long_answer_loss": 0.234,
-      "loss": 0.2136,
-      "short_answer_loss": NaN,
-      "step": 121,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.19,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4688166542224403e-05,
-      "long_answer_loss": 0.19,
-      "loss": 0.2076,
-      "short_answer_loss": NaN,
-      "step": 122,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.2186,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.4680817693084332e-05,
-      "long_answer_loss": 0.2186,
-      "loss": 0.2162,
-      "short_answer_loss": NaN,
-      "step": 123,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.234,
-      "grad_norm": 2.3125,
-      "learning_rate": 2.467338437539207e-05,
-      "long_answer_loss": 0.234,
-      "loss": 0.2225,
-      "short_answer_loss": NaN,
-      "step": 124,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.229,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4665866640694367e-05,
-      "long_answer_loss": 0.229,
-      "loss": 0.2066,
-      "short_answer_loss": NaN,
-      "step": 125,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.2273,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.4658264541123365e-05,
-      "long_answer_loss": 0.2273,
-      "loss": 0.2122,
-      "short_answer_loss": NaN,
-      "step": 126,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.2163,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4650578129396247e-05,
-      "long_answer_loss": 0.2163,
-      "loss": 0.2091,
-      "short_answer_loss": NaN,
-      "step": 127,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.2265,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.4642807458814864e-05,
-      "long_answer_loss": 0.2265,
-      "loss": 0.2126,
-      "short_answer_loss": NaN,
-      "step": 128,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1896,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.463495258326535e-05,
-      "long_answer_loss": 0.1896,
-      "loss": 0.2037,
-      "short_answer_loss": NaN,
-      "step": 129,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.2125,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4627013557217777e-05,
-      "long_answer_loss": 0.2125,
-      "loss": 0.1978,
-      "short_answer_loss": NaN,
-      "step": 130,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.2263,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.4618990435725758e-05,
-      "long_answer_loss": 0.2263,
-      "loss": 0.2285,
-      "short_answer_loss": NaN,
-      "step": 131,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1875,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.4610883274426076e-05,
-      "long_answer_loss": 0.1875,
-      "loss": 0.2163,
-      "short_answer_loss": NaN,
-      "step": 132,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1998,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.4602692129538286e-05,
-      "long_answer_loss": 0.1998,
-      "loss": 0.2166,
-      "short_answer_loss": NaN,
-      "step": 133,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2387,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.4594417057864327e-05,
-      "long_answer_loss": 0.2387,
-      "loss": 0.2039,
-      "short_answer_loss": NaN,
-      "step": 134,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2086,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.458605811678815e-05,
-      "long_answer_loss": 0.2086,
-      "loss": 0.2059,
-      "short_answer_loss": NaN,
-      "step": 135,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2005,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4577615364275292e-05,
-      "long_answer_loss": 0.2005,
-      "loss": 0.1979,
-      "short_answer_loss": NaN,
-      "step": 136,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2188,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.4569088858872478e-05,
-      "long_answer_loss": 0.2188,
-      "loss": 0.2078,
-      "short_answer_loss": NaN,
-      "step": 137,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2044,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.4560478659707236e-05,
-      "long_answer_loss": 0.2044,
-      "loss": 0.2228,
-      "short_answer_loss": NaN,
-      "step": 138,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1837,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4551784826487466e-05,
-      "long_answer_loss": 0.1837,
-      "loss": 0.2176,
-      "short_answer_loss": NaN,
-      "step": 139,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.2002,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.4543007419501034e-05,
-      "long_answer_loss": 0.2002,
-      "loss": 0.2067,
-      "short_answer_loss": NaN,
-      "step": 140,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.227,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.4534146499615356e-05,
-      "long_answer_loss": 0.227,
-      "loss": 0.2045,
-      "short_answer_loss": NaN,
-      "step": 141,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.2361,
-      "grad_norm": 1.875,
-      "learning_rate": 2.4525202128276962e-05,
-      "long_answer_loss": 0.2361,
-      "loss": 0.2197,
-      "short_answer_loss": NaN,
-      "step": 142,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1792,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.4516174367511095e-05,
-      "long_answer_loss": 0.1792,
-      "loss": 0.2087,
-      "short_answer_loss": NaN,
-      "step": 143,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.2059,
-      "grad_norm": 1.875,
-      "learning_rate": 2.450706327992126e-05,
-      "long_answer_loss": 0.2059,
-      "loss": 0.2137,
-      "short_answer_loss": NaN,
-      "step": 144,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.213,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.4497868928688794e-05,
-      "long_answer_loss": 0.213,
-      "loss": 0.2005,
-      "short_answer_loss": NaN,
-      "step": 145,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2122,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4488591377572434e-05,
-      "long_answer_loss": 0.2122,
-      "loss": 0.2188,
-      "short_answer_loss": NaN,
-      "step": 146,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2375,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.4479230690907868e-05,
-      "long_answer_loss": 0.2375,
-      "loss": 0.22,
-      "short_answer_loss": NaN,
-      "step": 147,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.163,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.4469786933607296e-05,
-      "long_answer_loss": 0.163,
-      "loss": 0.2075,
-      "short_answer_loss": NaN,
-      "step": 148,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2084,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.4460260171158973e-05,
-      "long_answer_loss": 0.2084,
-      "loss": 0.2139,
-      "short_answer_loss": NaN,
-      "step": 149,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2117,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.4450650469626758e-05,
-      "long_answer_loss": 0.2117,
-      "loss": 0.2037,
-      "short_answer_loss": NaN,
-      "step": 150,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.2042,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.4440957895649658e-05,
-      "long_answer_loss": 0.2042,
-      "loss": 0.209,
-      "short_answer_loss": NaN,
-      "step": 151,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.214,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.4431182516441363e-05,
-      "long_answer_loss": 0.214,
-      "loss": 0.2001,
-      "short_answer_loss": NaN,
-      "step": 152,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1751,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.4421324399789775e-05,
-      "long_answer_loss": 0.1751,
-      "loss": 0.218,
-      "short_answer_loss": NaN,
-      "step": 153,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1983,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.4411383614056554e-05,
-      "long_answer_loss": 0.1983,
-      "loss": 0.2138,
-      "short_answer_loss": NaN,
-      "step": 154,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.2409,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.440136022817662e-05,
-      "long_answer_loss": 0.2409,
-      "loss": 0.2068,
-      "short_answer_loss": NaN,
-      "step": 155,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.2302,
-      "grad_norm": 1.75,
-      "learning_rate": 2.4391254311657698e-05,
-      "long_answer_loss": 0.2302,
-      "loss": 0.2146,
-      "short_answer_loss": NaN,
-      "step": 156,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.2086,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.4381065934579827e-05,
-      "long_answer_loss": 0.2086,
-      "loss": 0.2047,
-      "short_answer_loss": NaN,
-      "step": 157,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.208,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.4370795167594864e-05,
-      "long_answer_loss": 0.208,
-      "loss": 0.2037,
-      "short_answer_loss": NaN,
-      "step": 158,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.212,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.4360442081926016e-05,
-      "long_answer_loss": 0.212,
-      "loss": 0.1997,
-      "short_answer_loss": NaN,
-      "step": 159,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.2122,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.435000674936732e-05,
-      "long_answer_loss": 0.2122,
-      "loss": 0.21,
-      "short_answer_loss": NaN,
-      "step": 160,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.2249,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.4339489242283166e-05,
-      "long_answer_loss": 0.2249,
-      "loss": 0.2093,
-      "short_answer_loss": NaN,
-      "step": 161,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.2184,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.4328889633607794e-05,
-      "long_answer_loss": 0.2184,
-      "loss": 0.2061,
-      "short_answer_loss": NaN,
-      "step": 162,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.2118,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.4318207996844767e-05,
-      "long_answer_loss": 0.2118,
-      "loss": 0.2087,
-      "short_answer_loss": NaN,
-      "step": 163,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1999,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.4307444406066488e-05,
-      "long_answer_loss": 0.1999,
-      "loss": 0.2065,
-      "short_answer_loss": NaN,
-      "step": 164,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1897,
-      "grad_norm": 1.6484375,
-      "learning_rate": 2.429659893591367e-05,
-      "long_answer_loss": 0.1897,
-      "loss": 0.2055,
-      "short_answer_loss": NaN,
-      "step": 165,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.2282,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.4285671661594827e-05,
-      "long_answer_loss": 0.2282,
-      "loss": 0.2235,
-      "short_answer_loss": NaN,
-      "step": 166,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.213,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.427466265888574e-05,
-      "long_answer_loss": 0.213,
-      "loss": 0.2049,
-      "short_answer_loss": NaN,
-      "step": 167,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.2374,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.426357200412895e-05,
-      "long_answer_loss": 0.2374,
-      "loss": 0.2186,
-      "short_answer_loss": NaN,
-      "step": 168,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1913,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.4252399774233216e-05,
-      "long_answer_loss": 0.1913,
-      "loss": 0.1997,
-      "short_answer_loss": NaN,
-      "step": 169,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.2179,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.4241146046672972e-05,
-      "long_answer_loss": 0.2179,
-      "loss": 0.2187,
-      "short_answer_loss": NaN,
-      "step": 170,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1981,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.4229810899487824e-05,
-      "long_answer_loss": 0.1981,
-      "loss": 0.2142,
-      "short_answer_loss": NaN,
-      "step": 171,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1987,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.421839441128197e-05,
-      "long_answer_loss": 0.1987,
-      "loss": 0.2143,
-      "short_answer_loss": NaN,
-      "step": 172,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1928,
-      "grad_norm": 2.328125,
-      "learning_rate": 2.4206896661223676e-05,
-      "long_answer_loss": 0.1928,
-      "loss": 0.2202,
-      "short_answer_loss": NaN,
-      "step": 173,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.2017,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4195317729044732e-05,
-      "long_answer_loss": 0.2017,
-      "loss": 0.2074,
-      "short_answer_loss": NaN,
-      "step": 174,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1975,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.4183657695039874e-05,
-      "long_answer_loss": 0.1975,
-      "loss": 0.2064,
-      "short_answer_loss": NaN,
-      "step": 175,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2251,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.417191664006625e-05,
-      "long_answer_loss": 0.2251,
-      "loss": 0.2147,
-      "short_answer_loss": NaN,
-      "step": 176,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2047,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.4160094645542857e-05,
-      "long_answer_loss": 0.2047,
-      "loss": 0.2036,
-      "short_answer_loss": NaN,
-      "step": 177,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2278,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.4148191793449974e-05,
-      "long_answer_loss": 0.2278,
-      "loss": 0.2078,
-      "short_answer_loss": NaN,
-      "step": 178,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2469,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4136208166328573e-05,
-      "long_answer_loss": 0.2469,
-      "loss": 0.2153,
-      "short_answer_loss": NaN,
-      "step": 179,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1945,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.4124143847279785e-05,
-      "long_answer_loss": 0.1945,
-      "loss": 0.1965,
-      "short_answer_loss": NaN,
-      "step": 180,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1612,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4111998919964297e-05,
-      "long_answer_loss": 0.1612,
-      "loss": 0.2107,
-      "short_answer_loss": NaN,
-      "step": 181,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2099,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.4099773468601773e-05,
-      "long_answer_loss": 0.2099,
-      "loss": 0.2054,
-      "short_answer_loss": NaN,
-      "step": 182,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.1939,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.408746757797028e-05,
-      "long_answer_loss": 0.1939,
-      "loss": 0.2171,
-      "short_answer_loss": NaN,
-      "step": 183,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2258,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4075081333405697e-05,
-      "long_answer_loss": 0.2258,
-      "loss": 0.21,
-      "short_answer_loss": NaN,
-      "step": 184,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.251,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.406261482080112e-05,
-      "long_answer_loss": 0.251,
-      "loss": 0.2233,
-      "short_answer_loss": NaN,
-      "step": 185,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.1729,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.4050068126606267e-05,
-      "long_answer_loss": 0.1729,
-      "loss": 0.2058,
-      "short_answer_loss": NaN,
-      "step": 186,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.1879,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.4037441337826884e-05,
-      "long_answer_loss": 0.1879,
-      "loss": 0.2084,
-      "short_answer_loss": NaN,
-      "step": 187,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1935,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.4024734542024135e-05,
-      "long_answer_loss": 0.1935,
-      "loss": 0.1992,
-      "short_answer_loss": NaN,
-      "step": 188,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.2111,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.401194782731399e-05,
-      "long_answer_loss": 0.2111,
-      "loss": 0.2078,
-      "short_answer_loss": NaN,
-      "step": 189,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.2188,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.3999081282366636e-05,
-      "long_answer_loss": 0.2188,
-      "loss": 0.2192,
-      "short_answer_loss": NaN,
-      "step": 190,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.2283,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.3986134996405832e-05,
-      "long_answer_loss": 0.2283,
-      "loss": 0.2084,
-      "short_answer_loss": NaN,
-      "step": 191,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1982,
-      "grad_norm": 1.75,
-      "learning_rate": 2.3973109059208315e-05,
-      "long_answer_loss": 0.1982,
-      "loss": 0.2095,
-      "short_answer_loss": NaN,
-      "step": 192,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.2264,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.396000356110316e-05,
-      "long_answer_loss": 0.2264,
-      "loss": 0.2161,
-      "short_answer_loss": NaN,
-      "step": 193,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1977,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.3946818592971176e-05,
-      "long_answer_loss": 0.1977,
-      "loss": 0.2139,
-      "short_answer_loss": NaN,
-      "step": 194,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2048,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.3933554246244243e-05,
-      "long_answer_loss": 0.2048,
-      "loss": 0.1998,
-      "short_answer_loss": NaN,
-      "step": 195,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1993,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.3920210612904715e-05,
-      "long_answer_loss": 0.1993,
-      "loss": 0.204,
-      "short_answer_loss": NaN,
-      "step": 196,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2245,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.3906787785484742e-05,
-      "long_answer_loss": 0.2245,
-      "loss": 0.1939,
-      "short_answer_loss": NaN,
-      "step": 197,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2195,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.3893285857065666e-05,
-      "long_answer_loss": 0.2195,
-      "loss": 0.2158,
-      "short_answer_loss": NaN,
-      "step": 198,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2314,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.3879704921277356e-05,
-      "long_answer_loss": 0.2314,
-      "loss": 0.2045,
-      "short_answer_loss": NaN,
-      "step": 199,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2089,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.386604507229756e-05,
-      "long_answer_loss": 0.2089,
-      "loss": 0.2042,
-      "short_answer_loss": NaN,
-      "step": 200,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1831,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.385230640485125e-05,
-      "long_answer_loss": 0.1831,
-      "loss": 0.1976,
-      "short_answer_loss": NaN,
-      "step": 201,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1918,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.383848901420998e-05,
-      "long_answer_loss": 0.1918,
-      "loss": 0.2033,
-      "short_answer_loss": NaN,
-      "step": 202,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1952,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.3824592996191204e-05,
-      "long_answer_loss": 0.1952,
-      "loss": 0.2051,
-      "short_answer_loss": NaN,
-      "step": 203,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.2222,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.3810618447157622e-05,
-      "long_answer_loss": 0.2222,
-      "loss": 0.2054,
-      "short_answer_loss": NaN,
-      "step": 204,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1614,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.3796565464016523e-05,
-      "long_answer_loss": 0.1614,
-      "loss": 0.2093,
-      "short_answer_loss": NaN,
-      "step": 205,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1581,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.378243414421909e-05,
-      "long_answer_loss": 0.1581,
-      "loss": 0.2025,
-      "short_answer_loss": NaN,
-      "step": 206,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.2477,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.3768224585759735e-05,
-      "long_answer_loss": 0.2477,
-      "loss": 0.2058,
-      "short_answer_loss": NaN,
-      "step": 207,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.2161,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.3753936887175433e-05,
-      "long_answer_loss": 0.2161,
-      "loss": 0.1956,
-      "short_answer_loss": NaN,
-      "step": 208,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1787,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.3739571147545007e-05,
-      "long_answer_loss": 0.1787,
-      "loss": 0.2096,
-      "short_answer_loss": NaN,
-      "step": 209,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1975,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.3725127466488483e-05,
-      "long_answer_loss": 0.1975,
-      "loss": 0.2039,
-      "short_answer_loss": NaN,
-      "step": 210,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.2476,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.371060594416636e-05,
-      "long_answer_loss": 0.2476,
-      "loss": 0.2051,
-      "short_answer_loss": NaN,
-      "step": 211,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1977,
-      "grad_norm": 1.875,
-      "learning_rate": 2.369600668127893e-05,
-      "long_answer_loss": 0.1977,
-      "loss": 0.1993,
-      "short_answer_loss": NaN,
-      "step": 212,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.198,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.3681329779065596e-05,
-      "long_answer_loss": 0.198,
-      "loss": 0.2029,
-      "short_answer_loss": NaN,
-      "step": 213,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.2019,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.366657533930414e-05,
-      "long_answer_loss": 0.2019,
-      "loss": 0.201,
-      "short_answer_loss": NaN,
-      "step": 214,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.2116,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.3651743464310038e-05,
-      "long_answer_loss": 0.2116,
-      "loss": 0.1928,
-      "short_answer_loss": NaN,
-      "step": 215,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.2155,
-      "grad_norm": 2.125,
-      "learning_rate": 2.3636834256935745e-05,
-      "long_answer_loss": 0.2155,
-      "loss": 0.2115,
-      "short_answer_loss": NaN,
-      "step": 216,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1885,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.3621847820569988e-05,
-      "long_answer_loss": 0.1885,
-      "loss": 0.2007,
-      "short_answer_loss": NaN,
-      "step": 217,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1827,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.3606784259137033e-05,
-      "long_answer_loss": 0.1827,
-      "loss": 0.1991,
-      "short_answer_loss": NaN,
-      "step": 218,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.2293,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.3591643677095973e-05,
-      "long_answer_loss": 0.2293,
-      "loss": 0.2025,
-      "short_answer_loss": NaN,
-      "step": 219,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1983,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.3576426179440014e-05,
-      "long_answer_loss": 0.1983,
-      "loss": 0.2017,
-      "short_answer_loss": NaN,
-      "step": 220,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.2172,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.3561131871695736e-05,
-      "long_answer_loss": 0.2172,
-      "loss": 0.2012,
-      "short_answer_loss": NaN,
-      "step": 221,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1724,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.3545760859922354e-05,
-      "long_answer_loss": 0.1724,
-      "loss": 0.2022,
-      "short_answer_loss": NaN,
-      "step": 222,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1781,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.3530313250710998e-05,
-      "long_answer_loss": 0.1781,
-      "loss": 0.2049,
-      "short_answer_loss": NaN,
-      "step": 223,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1954,
-      "grad_norm": 1.875,
-      "learning_rate": 2.351478915118397e-05,
-      "long_answer_loss": 0.1954,
-      "loss": 0.1994,
-      "short_answer_loss": NaN,
-      "step": 224,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.2026,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.349918866899399e-05,
-      "long_answer_loss": 0.2026,
-      "loss": 0.2201,
-      "short_answer_loss": NaN,
-      "step": 225,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.2047,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.348351191232346e-05,
-      "long_answer_loss": 0.2047,
-      "loss": 0.1915,
-      "short_answer_loss": NaN,
-      "step": 226,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1743,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.346775898988372e-05,
-      "long_answer_loss": 0.1743,
-      "loss": 0.2042,
-      "short_answer_loss": NaN,
-      "step": 227,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1933,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.345193001091428e-05,
-      "long_answer_loss": 0.1933,
-      "loss": 0.2044,
-      "short_answer_loss": NaN,
-      "step": 228,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1829,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.3436025085182064e-05,
-      "long_answer_loss": 0.1829,
-      "loss": 0.1989,
-      "short_answer_loss": NaN,
-      "step": 229,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.187,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.3420044322980662e-05,
-      "long_answer_loss": 0.187,
-      "loss": 0.2025,
-      "short_answer_loss": NaN,
-      "step": 230,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1994,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.340398783512954e-05,
-      "long_answer_loss": 0.1994,
-      "loss": 0.212,
-      "short_answer_loss": NaN,
-      "step": 231,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.2226,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.3387855732973307e-05,
-      "long_answer_loss": 0.2226,
-      "loss": 0.2044,
-      "short_answer_loss": NaN,
-      "step": 232,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1717,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.3371648128380918e-05,
-      "long_answer_loss": 0.1717,
-      "loss": 0.2019,
-      "short_answer_loss": NaN,
-      "step": 233,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.2201,
-      "grad_norm": 2.0,
-      "learning_rate": 2.3355365133744894e-05,
-      "long_answer_loss": 0.2201,
-      "loss": 0.2004,
-      "short_answer_loss": NaN,
-      "step": 234,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.166,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.3339006861980562e-05,
-      "long_answer_loss": 0.166,
-      "loss": 0.2043,
-      "short_answer_loss": NaN,
-      "step": 235,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.2252,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.3322573426525262e-05,
-      "long_answer_loss": 0.2252,
-      "loss": 0.2121,
-      "short_answer_loss": NaN,
-      "step": 236,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.2206,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.330606494133756e-05,
-      "long_answer_loss": 0.2206,
-      "loss": 0.216,
-      "short_answer_loss": NaN,
-      "step": 237,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.2155,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.328948152089645e-05,
-      "long_answer_loss": 0.2155,
-      "loss": 0.2058,
-      "short_answer_loss": NaN,
-      "step": 238,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.228,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.327282328020058e-05,
-      "long_answer_loss": 0.228,
-      "loss": 0.2064,
-      "short_answer_loss": NaN,
-      "step": 239,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1713,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.3256090334767443e-05,
-      "long_answer_loss": 0.1713,
-      "loss": 0.1939,
-      "short_answer_loss": NaN,
-      "step": 240,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1933,
-      "grad_norm": 1.6484375,
-      "learning_rate": 2.3239282800632564e-05,
-      "long_answer_loss": 0.1933,
-      "loss": 0.2032,
-      "short_answer_loss": NaN,
-      "step": 241,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1856,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.322240079434872e-05,
-      "long_answer_loss": 0.1856,
-      "loss": 0.2067,
-      "short_answer_loss": NaN,
-      "step": 242,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1947,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.320544443298512e-05,
-      "long_answer_loss": 0.1947,
-      "loss": 0.2113,
-      "short_answer_loss": NaN,
-      "step": 243,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1911,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.3188413834126573e-05,
-      "long_answer_loss": 0.1911,
-      "loss": 0.202,
-      "short_answer_loss": NaN,
-      "step": 244,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.2166,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.317130911587272e-05,
-      "long_answer_loss": 0.2166,
-      "loss": 0.2018,
-      "short_answer_loss": NaN,
-      "step": 245,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.2132,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.3154130396837166e-05,
-      "long_answer_loss": 0.2132,
-      "loss": 0.1956,
-      "short_answer_loss": NaN,
-      "step": 246,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1885,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.313687779614669e-05,
-      "long_answer_loss": 0.1885,
-      "loss": 0.2011,
-      "short_answer_loss": NaN,
-      "step": 247,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.2018,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.31195514334404e-05,
-      "long_answer_loss": 0.2018,
-      "loss": 0.2032,
-      "short_answer_loss": NaN,
-      "step": 248,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.2596,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.3102151428868912e-05,
-      "long_answer_loss": 0.2596,
-      "loss": 0.2129,
-      "short_answer_loss": NaN,
-      "step": 249,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1831,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.3084677903093528e-05,
-      "long_answer_loss": 0.1831,
-      "loss": 0.1961,
-      "short_answer_loss": NaN,
-      "step": 250,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.199,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.306713097728536e-05,
-      "long_answer_loss": 0.199,
-      "loss": 0.1935,
-      "short_answer_loss": NaN,
-      "step": 251,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.2525,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.3049510773124546e-05,
-      "long_answer_loss": 0.2525,
-      "loss": 0.2168,
-      "short_answer_loss": NaN,
-      "step": 252,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.226,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.303181741279936e-05,
-      "long_answer_loss": 0.226,
-      "loss": 0.1991,
-      "short_answer_loss": NaN,
-      "step": 253,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1931,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.3014051019005383e-05,
-      "long_answer_loss": 0.1931,
-      "loss": 0.2099,
-      "short_answer_loss": NaN,
-      "step": 254,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.2221,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.2996211714944653e-05,
-      "long_answer_loss": 0.2221,
-      "loss": 0.2075,
-      "short_answer_loss": NaN,
-      "step": 255,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.2067,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.297829962432481e-05,
-      "long_answer_loss": 0.2067,
-      "loss": 0.1982,
-      "short_answer_loss": NaN,
-      "step": 256,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.234,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.296031487135824e-05,
-      "long_answer_loss": 0.234,
-      "loss": 0.2171,
-      "short_answer_loss": NaN,
-      "step": 257,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.2223,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.294225758076119e-05,
-      "long_answer_loss": 0.2223,
-      "loss": 0.2007,
-      "short_answer_loss": NaN,
-      "step": 258,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1731,
-      "grad_norm": 1.75,
-      "learning_rate": 2.292412787775295e-05,
-      "long_answer_loss": 0.1731,
-      "loss": 0.1939,
-      "short_answer_loss": NaN,
-      "step": 259,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1604,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.290592588805494e-05,
-      "long_answer_loss": 0.1604,
-      "loss": 0.1877,
-      "short_answer_loss": NaN,
-      "step": 260,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.2115,
-      "grad_norm": 2.0,
-      "learning_rate": 2.2887651737889866e-05,
-      "long_answer_loss": 0.2115,
-      "loss": 0.2019,
-      "short_answer_loss": NaN,
-      "step": 261,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1951,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.2869305553980823e-05,
-      "long_answer_loss": 0.1951,
-      "loss": 0.2116,
-      "short_answer_loss": NaN,
-      "step": 262,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.2196,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.2850887463550442e-05,
-      "long_answer_loss": 0.2196,
-      "loss": 0.1976,
-      "short_answer_loss": NaN,
-      "step": 263,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.2057,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.2832397594319983e-05,
-      "long_answer_loss": 0.2057,
-      "loss": 0.2028,
-      "short_answer_loss": NaN,
-      "step": 264,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.212,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.2813836074508467e-05,
-      "long_answer_loss": 0.212,
-      "loss": 0.2052,
-      "short_answer_loss": NaN,
-      "step": 265,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.2168,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.2795203032831776e-05,
-      "long_answer_loss": 0.2168,
-      "loss": 0.1947,
-      "short_answer_loss": NaN,
-      "step": 266,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1731,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.2776498598501767e-05,
-      "long_answer_loss": 0.1731,
-      "loss": 0.1975,
-      "short_answer_loss": NaN,
-      "step": 267,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1762,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.2757722901225367e-05,
-      "long_answer_loss": 0.1762,
-      "loss": 0.2008,
-      "short_answer_loss": NaN,
-      "step": 268,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.2079,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.2738876071203688e-05,
-      "long_answer_loss": 0.2079,
-      "loss": 0.2111,
-      "short_answer_loss": NaN,
-      "step": 269,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.2049,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.271995823913111e-05,
-      "long_answer_loss": 0.2049,
-      "loss": 0.2021,
-      "short_answer_loss": NaN,
-      "step": 270,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1998,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.270096953619439e-05,
-      "long_answer_loss": 0.1998,
-      "loss": 0.2074,
-      "short_answer_loss": NaN,
-      "step": 271,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1979,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.2681910094071724e-05,
-      "long_answer_loss": 0.1979,
-      "loss": 0.1884,
-      "short_answer_loss": NaN,
-      "step": 272,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1774,
-      "grad_norm": 1.7109375,
-      "learning_rate": 2.2662780044931874e-05,
-      "long_answer_loss": 0.1774,
-      "loss": 0.1848,
-      "short_answer_loss": NaN,
-      "step": 273,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.2164,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.264357952143322e-05,
-      "long_answer_loss": 0.2164,
-      "loss": 0.2013,
-      "short_answer_loss": NaN,
-      "step": 274,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.2208,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.2624308656722846e-05,
-      "long_answer_loss": 0.2208,
-      "loss": 0.1935,
-      "short_answer_loss": NaN,
-      "step": 275,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.2249,
-      "grad_norm": 2.0,
-      "learning_rate": 2.260496758443563e-05,
-      "long_answer_loss": 0.2249,
-      "loss": 0.2183,
-      "short_answer_loss": NaN,
-      "step": 276,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.2462,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.258555643869331e-05,
-      "long_answer_loss": 0.2462,
-      "loss": 0.2003,
-      "short_answer_loss": NaN,
-      "step": 277,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.2265,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.256607535410354e-05,
-      "long_answer_loss": 0.2265,
-      "loss": 0.2058,
-      "short_answer_loss": NaN,
-      "step": 278,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.185,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.2546524465758973e-05,
-      "long_answer_loss": 0.185,
-      "loss": 0.1978,
-      "short_answer_loss": NaN,
-      "step": 279,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.1871,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.252690390923633e-05,
-      "long_answer_loss": 0.1871,
-      "loss": 0.2059,
-      "short_answer_loss": NaN,
-      "step": 280,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.2192,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.2507213820595435e-05,
-      "long_answer_loss": 0.2192,
-      "loss": 0.2039,
-      "short_answer_loss": NaN,
-      "step": 281,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1982,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.2487454336378303e-05,
-      "long_answer_loss": 0.1982,
-      "loss": 0.1963,
-      "short_answer_loss": NaN,
-      "step": 282,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.2183,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.246762559360816e-05,
-      "long_answer_loss": 0.2183,
-      "loss": 0.1968,
-      "short_answer_loss": NaN,
-      "step": 283,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.2063,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.244772772978852e-05,
-      "long_answer_loss": 0.2063,
-      "loss": 0.2033,
-      "short_answer_loss": NaN,
-      "step": 284,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1825,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.2427760882902217e-05,
-      "long_answer_loss": 0.1825,
-      "loss": 0.2233,
-      "short_answer_loss": NaN,
-      "step": 285,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1893,
-      "grad_norm": 1.75,
-      "learning_rate": 2.2407725191410446e-05,
-      "long_answer_loss": 0.1893,
-      "loss": 0.1977,
-      "short_answer_loss": NaN,
-      "step": 286,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.201,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.2387620794251824e-05,
-      "long_answer_loss": 0.201,
-      "loss": 0.207,
-      "short_answer_loss": NaN,
-      "step": 287,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.2089,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.2367447830841398e-05,
-      "long_answer_loss": 0.2089,
-      "loss": 0.1944,
-      "short_answer_loss": NaN,
-      "step": 288,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.2115,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.234720644106969e-05,
-      "long_answer_loss": 0.2115,
-      "loss": 0.1969,
-      "short_answer_loss": NaN,
-      "step": 289,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1723,
-      "grad_norm": 1.609375,
-      "learning_rate": 2.2326896765301746e-05,
-      "long_answer_loss": 0.1723,
-      "loss": 0.1827,
-      "short_answer_loss": NaN,
-      "step": 290,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.2066,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.2306518944376125e-05,
-      "long_answer_loss": 0.2066,
-      "loss": 0.2062,
-      "short_answer_loss": NaN,
-      "step": 291,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1793,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.2286073119603952e-05,
-      "long_answer_loss": 0.1793,
-      "loss": 0.1906,
-      "short_answer_loss": NaN,
-      "step": 292,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.2182,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.2265559432767924e-05,
-      "long_answer_loss": 0.2182,
-      "loss": 0.2009,
-      "short_answer_loss": NaN,
-      "step": 293,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1888,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.224497802612134e-05,
-      "long_answer_loss": 0.1888,
-      "loss": 0.1953,
-      "short_answer_loss": NaN,
-      "step": 294,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1951,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.2224329042387093e-05,
-      "long_answer_loss": 0.1951,
-      "loss": 0.2054,
-      "short_answer_loss": NaN,
-      "step": 295,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1802,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.2203612624756704e-05,
-      "long_answer_loss": 0.1802,
-      "loss": 0.1995,
-      "short_answer_loss": NaN,
-      "step": 296,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.2195,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.218282891688931e-05,
-      "long_answer_loss": 0.2195,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 297,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1794,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.216197806291068e-05,
-      "long_answer_loss": 0.1794,
-      "loss": 0.1977,
-      "short_answer_loss": NaN,
-      "step": 298,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.199,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.2141060207412224e-05,
-      "long_answer_loss": 0.199,
-      "loss": 0.2008,
-      "short_answer_loss": NaN,
-      "step": 299,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1984,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.2120075495449944e-05,
-      "long_answer_loss": 0.1984,
-      "loss": 0.1934,
-      "short_answer_loss": NaN,
-      "step": 300,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.2078,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.2099024072543495e-05,
-      "long_answer_loss": 0.2078,
-      "loss": 0.2045,
-      "short_answer_loss": NaN,
-      "step": 301,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1886,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.2077906084675126e-05,
-      "long_answer_loss": 0.1886,
-      "loss": 0.1967,
-      "short_answer_loss": NaN,
-      "step": 302,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1978,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.2056721678288693e-05,
-      "long_answer_loss": 0.1978,
-      "loss": 0.2,
-      "short_answer_loss": NaN,
-      "step": 303,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.199,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.2035471000288628e-05,
-      "long_answer_loss": 0.199,
-      "loss": 0.2,
-      "short_answer_loss": NaN,
-      "step": 304,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.2371,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.201415419803893e-05,
-      "long_answer_loss": 0.2371,
-      "loss": 0.2014,
-      "short_answer_loss": NaN,
-      "step": 305,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1504,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.199277141936214e-05,
-      "long_answer_loss": 0.1504,
-      "loss": 0.2031,
-      "short_answer_loss": NaN,
-      "step": 306,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1945,
-      "grad_norm": 1.6953125,
-      "learning_rate": 2.1971322812538314e-05,
-      "long_answer_loss": 0.1945,
-      "loss": 0.1907,
-      "short_answer_loss": NaN,
-      "step": 307,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1924,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.1949808526304006e-05,
-      "long_answer_loss": 0.1924,
-      "loss": 0.2033,
-      "short_answer_loss": NaN,
-      "step": 308,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1645,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.1928228709851212e-05,
-      "long_answer_loss": 0.1645,
-      "loss": 0.1964,
-      "short_answer_loss": NaN,
-      "step": 309,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1908,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.190658351282637e-05,
-      "long_answer_loss": 0.1908,
-      "loss": 0.1965,
-      "short_answer_loss": NaN,
-      "step": 310,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.2053,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.1884873085329276e-05,
-      "long_answer_loss": 0.2053,
-      "loss": 0.1976,
-      "short_answer_loss": NaN,
-      "step": 311,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1693,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.1863097577912107e-05,
-      "long_answer_loss": 0.1693,
-      "loss": 0.2003,
-      "short_answer_loss": NaN,
-      "step": 312,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.191,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.1841257141578304e-05,
-      "long_answer_loss": 0.191,
-      "loss": 0.2086,
-      "short_answer_loss": NaN,
-      "step": 313,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.217,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.181935192778159e-05,
-      "long_answer_loss": 0.217,
-      "loss": 0.205,
-      "short_answer_loss": NaN,
-      "step": 314,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1765,
-      "grad_norm": 1.6171875,
-      "learning_rate": 2.1797382088424866e-05,
-      "long_answer_loss": 0.1765,
-      "loss": 0.1931,
-      "short_answer_loss": NaN,
-      "step": 315,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.242,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1775347775859205e-05,
-      "long_answer_loss": 0.242,
-      "loss": 0.2047,
-      "short_answer_loss": NaN,
-      "step": 316,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.2099,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.175324914288276e-05,
-      "long_answer_loss": 0.2099,
-      "loss": 0.1956,
-      "short_answer_loss": NaN,
-      "step": 317,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1706,
-      "grad_norm": 1.75,
-      "learning_rate": 2.173108634273972e-05,
-      "long_answer_loss": 0.1706,
-      "loss": 0.2052,
-      "short_answer_loss": NaN,
-      "step": 318,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1948,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.1708859529119242e-05,
-      "long_answer_loss": 0.1948,
-      "loss": 0.1881,
-      "short_answer_loss": NaN,
-      "step": 319,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.2132,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.1686568856154397e-05,
-      "long_answer_loss": 0.2132,
-      "loss": 0.1945,
-      "short_answer_loss": NaN,
-      "step": 320,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.2051,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.166421447842108e-05,
-      "long_answer_loss": 0.2051,
-      "loss": 0.2039,
-      "short_answer_loss": NaN,
-      "step": 321,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.2143,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.1641796550936964e-05,
-      "long_answer_loss": 0.2143,
-      "loss": 0.2026,
-      "short_answer_loss": NaN,
-      "step": 322,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.2044,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.1619315229160396e-05,
-      "long_answer_loss": 0.2044,
-      "loss": 0.2007,
-      "short_answer_loss": NaN,
-      "step": 323,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.2178,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.1596770668989347e-05,
-      "long_answer_loss": 0.2178,
-      "loss": 0.2133,
-      "short_answer_loss": NaN,
-      "step": 324,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1888,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.1574163026760308e-05,
-      "long_answer_loss": 0.1888,
-      "loss": 0.1957,
-      "short_answer_loss": NaN,
-      "step": 325,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1592,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.1551492459247227e-05,
-      "long_answer_loss": 0.1592,
-      "loss": 0.2062,
-      "short_answer_loss": NaN,
-      "step": 326,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1695,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.1528759123660404e-05,
-      "long_answer_loss": 0.1695,
-      "loss": 0.198,
-      "short_answer_loss": NaN,
-      "step": 327,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1931,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.1505963177645404e-05,
-      "long_answer_loss": 0.1931,
-      "loss": 0.1908,
-      "short_answer_loss": NaN,
-      "step": 328,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1755,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.1483104779281975e-05,
-      "long_answer_loss": 0.1755,
-      "loss": 0.205,
-      "short_answer_loss": NaN,
-      "step": 329,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.227,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1460184087082944e-05,
-      "long_answer_loss": 0.227,
-      "loss": 0.2101,
-      "short_answer_loss": NaN,
-      "step": 330,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1985,
-      "grad_norm": 1.6640625,
-      "learning_rate": 2.1437201259993112e-05,
-      "long_answer_loss": 0.1985,
-      "loss": 0.2003,
-      "short_answer_loss": NaN,
-      "step": 331,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1934,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.141415645738816e-05,
-      "long_answer_loss": 0.1934,
-      "loss": 0.201,
-      "short_answer_loss": NaN,
-      "step": 332,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.2029,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1391049839073544e-05,
-      "long_answer_loss": 0.2029,
-      "loss": 0.1983,
-      "short_answer_loss": NaN,
-      "step": 333,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1792,
-      "grad_norm": 1.6953125,
-      "learning_rate": 2.136788156528339e-05,
-      "long_answer_loss": 0.1792,
-      "loss": 0.1974,
-      "short_answer_loss": NaN,
-      "step": 334,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1899,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.134465179667936e-05,
-      "long_answer_loss": 0.1899,
-      "loss": 0.1941,
-      "short_answer_loss": NaN,
-      "step": 335,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1943,
-      "grad_norm": 1.7109375,
-      "learning_rate": 2.1321360694349573e-05,
-      "long_answer_loss": 0.1943,
-      "loss": 0.2083,
-      "short_answer_loss": NaN,
-      "step": 336,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1913,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.129800841980746e-05,
-      "long_answer_loss": 0.1913,
-      "loss": 0.1945,
-      "short_answer_loss": NaN,
-      "step": 337,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.2173,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.1274595134990656e-05,
-      "long_answer_loss": 0.2173,
-      "loss": 0.1996,
-      "short_answer_loss": NaN,
-      "step": 338,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1984,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.1251121002259875e-05,
-      "long_answer_loss": 0.1984,
-      "loss": 0.2004,
-      "short_answer_loss": NaN,
-      "step": 339,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.2009,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.122758618439779e-05,
-      "long_answer_loss": 0.2009,
-      "loss": 0.2014,
-      "short_answer_loss": NaN,
-      "step": 340,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1869,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.120399084460789e-05,
-      "long_answer_loss": 0.1869,
-      "loss": 0.1843,
-      "short_answer_loss": NaN,
-      "step": 341,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1868,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.1180335146513363e-05,
-      "long_answer_loss": 0.1868,
-      "loss": 0.1896,
-      "short_answer_loss": NaN,
-      "step": 342,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1844,
-      "grad_norm": 1.75,
-      "learning_rate": 2.1156619254155948e-05,
-      "long_answer_loss": 0.1844,
-      "loss": 0.1888,
-      "short_answer_loss": NaN,
-      "step": 343,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.2027,
-      "grad_norm": 1.6953125,
-      "learning_rate": 2.1132843331994817e-05,
-      "long_answer_loss": 0.2027,
-      "loss": 0.1868,
-      "short_answer_loss": NaN,
-      "step": 344,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1783,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.1109007544905402e-05,
-      "long_answer_loss": 0.1783,
-      "loss": 0.2011,
-      "short_answer_loss": NaN,
-      "step": 345,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.2077,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.108511205817829e-05,
-      "long_answer_loss": 0.2077,
-      "loss": 0.198,
-      "short_answer_loss": NaN,
-      "step": 346,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1811,
-      "grad_norm": 1.671875,
-      "learning_rate": 2.1061157037518057e-05,
-      "long_answer_loss": 0.1811,
-      "loss": 0.1946,
-      "short_answer_loss": NaN,
-      "step": 347,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.173,
-      "grad_norm": 1.75,
-      "learning_rate": 2.1037142649042107e-05,
-      "long_answer_loss": 0.173,
-      "loss": 0.1922,
-      "short_answer_loss": NaN,
-      "step": 348,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1804,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1013069059279542e-05,
-      "long_answer_loss": 0.1804,
-      "loss": 0.1918,
-      "short_answer_loss": NaN,
-      "step": 349,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1974,
-      "grad_norm": 1.65625,
-      "learning_rate": 2.0988936435170004e-05,
-      "long_answer_loss": 0.1974,
-      "loss": 0.1914,
-      "short_answer_loss": NaN,
-      "step": 350,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.2522,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.09647449440625e-05,
-      "long_answer_loss": 0.2522,
-      "loss": 0.2157,
-      "short_answer_loss": NaN,
-      "step": 351,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.2153,
-      "grad_norm": 1.75,
-      "learning_rate": 2.094049475371426e-05,
-      "long_answer_loss": 0.2153,
-      "loss": 0.2018,
-      "short_answer_loss": NaN,
-      "step": 352,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1612,
-      "grad_norm": 1.609375,
-      "learning_rate": 2.091618603228957e-05,
-      "long_answer_loss": 0.1612,
-      "loss": 0.1818,
-      "short_answer_loss": NaN,
-      "step": 353,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2161,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.0891818948358597e-05,
-      "long_answer_loss": 0.2161,
-      "loss": 0.1963,
-      "short_answer_loss": NaN,
-      "step": 354,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2255,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.086739367089623e-05,
-      "long_answer_loss": 0.2255,
-      "loss": 0.2045,
-      "short_answer_loss": NaN,
-      "step": 355,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1808,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.08429103692809e-05,
-      "long_answer_loss": 0.1808,
-      "loss": 0.1989,
-      "short_answer_loss": NaN,
-      "step": 356,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2142,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.0818369213293422e-05,
-      "long_answer_loss": 0.2142,
-      "loss": 0.2053,
-      "short_answer_loss": NaN,
-      "step": 357,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2077,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.0793770373115782e-05,
-      "long_answer_loss": 0.2077,
-      "loss": 0.1966,
-      "short_answer_loss": NaN,
-      "step": 358,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2175,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.0769114019330006e-05,
-      "long_answer_loss": 0.2175,
-      "loss": 0.2082,
-      "short_answer_loss": NaN,
-      "step": 359,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.2193,
-      "grad_norm": 1.75,
-      "learning_rate": 2.074440032291693e-05,
-      "long_answer_loss": 0.2193,
-      "loss": 0.1979,
-      "short_answer_loss": NaN,
-      "step": 360,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1886,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.0719629455255052e-05,
-      "long_answer_loss": 0.1886,
-      "loss": 0.1982,
-      "short_answer_loss": NaN,
-      "step": 361,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1853,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.0694801588119315e-05,
-      "long_answer_loss": 0.1853,
-      "loss": 0.199,
-      "short_answer_loss": NaN,
-      "step": 362,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.2133,
-      "grad_norm": 1.671875,
-      "learning_rate": 2.066991689367994e-05,
-      "long_answer_loss": 0.2133,
-      "loss": 0.1855,
-      "short_answer_loss": NaN,
-      "step": 363,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1858,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.0644975544501206e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1977,
-      "short_answer_loss": NaN,
-      "step": 364,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.182,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.061997771354028e-05,
-      "long_answer_loss": 0.182,
-      "loss": 0.1867,
-      "short_answer_loss": NaN,
-      "step": 365,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.2098,
-      "grad_norm": 1.75,
-      "learning_rate": 2.0594923574145994e-05,
-      "long_answer_loss": 0.2098,
-      "loss": 0.1901,
-      "short_answer_loss": NaN,
-      "step": 366,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1777,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.0569813300057667e-05,
-      "long_answer_loss": 0.1777,
-      "loss": 0.1942,
-      "short_answer_loss": NaN,
-      "step": 367,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1621,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.054464706540387e-05,
-      "long_answer_loss": 0.1621,
-      "loss": 0.1891,
-      "short_answer_loss": NaN,
-      "step": 368,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.2067,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.0519425044701256e-05,
-      "long_answer_loss": 0.2067,
-      "loss": 0.1956,
-      "short_answer_loss": NaN,
-      "step": 369,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.2081,
-      "grad_norm": 1.875,
-      "learning_rate": 2.049414741285331e-05,
-      "long_answer_loss": 0.2081,
-      "loss": 0.2041,
-      "short_answer_loss": NaN,
-      "step": 370,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1938,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.0468814345149173e-05,
-      "long_answer_loss": 0.1938,
-      "loss": 0.1946,
-      "short_answer_loss": NaN,
-      "step": 371,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1969,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.0443426017262395e-05,
-      "long_answer_loss": 0.1969,
-      "loss": 0.1923,
-      "short_answer_loss": NaN,
-      "step": 372,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1651,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.0417982605249734e-05,
-      "long_answer_loss": 0.1651,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 373,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1948,
-      "grad_norm": 1.6328125,
-      "learning_rate": 2.0392484285549936e-05,
-      "long_answer_loss": 0.1948,
-      "loss": 0.1997,
-      "short_answer_loss": NaN,
-      "step": 374,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.2046,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.0366931234982513e-05,
-      "long_answer_loss": 0.2046,
-      "loss": 0.2086,
-      "short_answer_loss": NaN,
-      "step": 375,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.2052,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.034132363074649e-05,
-      "long_answer_loss": 0.2052,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 376,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.2002,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.0315661650419222e-05,
-      "long_answer_loss": 0.2002,
-      "loss": 0.1858,
-      "short_answer_loss": NaN,
-      "step": 377,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1988,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.028994547195512e-05,
-      "long_answer_loss": 0.1988,
-      "loss": 0.1907,
-      "short_answer_loss": NaN,
-      "step": 378,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.2357,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.026417527368445e-05,
-      "long_answer_loss": 0.2357,
-      "loss": 0.219,
-      "short_answer_loss": NaN,
-      "step": 379,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.2111,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.0238351234312063e-05,
-      "long_answer_loss": 0.2111,
-      "loss": 0.1933,
-      "short_answer_loss": NaN,
-      "step": 380,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.2037,
-      "grad_norm": 1.6171875,
-      "learning_rate": 2.0212473532916192e-05,
-      "long_answer_loss": 0.2037,
-      "loss": 0.1859,
-      "short_answer_loss": NaN,
-      "step": 381,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.2295,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.0186542348947185e-05,
-      "long_answer_loss": 0.2295,
-      "loss": 0.2089,
-      "short_answer_loss": NaN,
-      "step": 382,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1794,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.0160557862226266e-05,
-      "long_answer_loss": 0.1794,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 383,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1677,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.013452025294429e-05,
-      "long_answer_loss": 0.1677,
-      "loss": 0.1936,
-      "short_answer_loss": NaN,
-      "step": 384,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.212,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.0108429701660496e-05,
-      "long_answer_loss": 0.212,
-      "loss": 0.196,
-      "short_answer_loss": NaN,
-      "step": 385,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1945,
-      "grad_norm": 1.625,
-      "learning_rate": 2.008228638930125e-05,
-      "long_answer_loss": 0.1945,
-      "loss": 0.1926,
-      "short_answer_loss": NaN,
-      "step": 386,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.201,
-      "grad_norm": 1.75,
-      "learning_rate": 2.0056090497158797e-05,
-      "long_answer_loss": 0.201,
-      "loss": 0.1863,
-      "short_answer_loss": NaN,
-      "step": 387,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.2279,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.002984220688999e-05,
-      "long_answer_loss": 0.2279,
-      "loss": 0.2077,
-      "short_answer_loss": NaN,
-      "step": 388,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1748,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.0003541700515056e-05,
-      "long_answer_loss": 0.1748,
-      "loss": 0.1907,
-      "short_answer_loss": NaN,
-      "step": 389,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1925,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.9977189160416293e-05,
-      "long_answer_loss": 0.1925,
-      "loss": 0.1977,
-      "short_answer_loss": NaN,
-      "step": 390,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.205,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.9950784769336856e-05,
-      "long_answer_loss": 0.205,
-      "loss": 0.1951,
-      "short_answer_loss": NaN,
-      "step": 391,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.2024,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.9924328710379443e-05,
-      "long_answer_loss": 0.2024,
-      "loss": 0.2014,
-      "short_answer_loss": NaN,
-      "step": 392,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.2031,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.989782116700506e-05,
-      "long_answer_loss": 0.2031,
-      "loss": 0.1955,
-      "short_answer_loss": NaN,
-      "step": 393,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1843,
-      "grad_norm": 1.8671875,
-      "learning_rate": 1.9871262323031725e-05,
-      "long_answer_loss": 0.1843,
-      "loss": 0.2188,
-      "short_answer_loss": NaN,
-      "step": 394,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.2244,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.9844652362633214e-05,
-      "long_answer_loss": 0.2244,
-      "loss": 0.2111,
-      "short_answer_loss": NaN,
-      "step": 395,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.2076,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.9817991470337756e-05,
-      "long_answer_loss": 0.2076,
-      "loss": 0.1906,
-      "short_answer_loss": NaN,
-      "step": 396,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.182,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.9791279831026783e-05,
-      "long_answer_loss": 0.182,
-      "loss": 0.1974,
-      "short_answer_loss": NaN,
-      "step": 397,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1739,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.9764517629933632e-05,
-      "long_answer_loss": 0.1739,
-      "loss": 0.1918,
-      "short_answer_loss": NaN,
-      "step": 398,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.2019,
-      "grad_norm": 1.859375,
-      "learning_rate": 1.9737705052642257e-05,
-      "long_answer_loss": 0.2019,
-      "loss": 0.1957,
-      "short_answer_loss": NaN,
-      "step": 399,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1593,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.9710842285085963e-05,
-      "long_answer_loss": 0.1593,
-      "loss": 0.1871,
-      "short_answer_loss": NaN,
-      "step": 400,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.2025,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.968392951354609e-05,
-      "long_answer_loss": 0.2025,
-      "loss": 0.1932,
-      "short_answer_loss": NaN,
-      "step": 401,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.2095,
-      "grad_norm": 2.0,
-      "learning_rate": 1.965696692465074e-05,
-      "long_answer_loss": 0.2095,
-      "loss": 0.1955,
-      "short_answer_loss": NaN,
-      "step": 402,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.2044,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.962995470537346e-05,
-      "long_answer_loss": 0.2044,
-      "loss": 0.1873,
-      "short_answer_loss": NaN,
-      "step": 403,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.2193,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.960289304303199e-05,
-      "long_answer_loss": 0.2193,
-      "loss": 0.1903,
-      "short_answer_loss": NaN,
-      "step": 404,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1501,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.9575782125286907e-05,
-      "long_answer_loss": 0.1501,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 405,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1722,
-      "grad_norm": 1.9609375,
-      "learning_rate": 1.954862214014038e-05,
-      "long_answer_loss": 0.1722,
-      "loss": 0.1893,
-      "short_answer_loss": NaN,
-      "step": 406,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1493,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.952141327593481e-05,
-      "long_answer_loss": 0.1493,
-      "loss": 0.1941,
-      "short_answer_loss": NaN,
-      "step": 407,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1972,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.949415572135158e-05,
-      "long_answer_loss": 0.1972,
-      "loss": 0.2043,
-      "short_answer_loss": NaN,
-      "step": 408,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1868,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.9466849665409695e-05,
-      "long_answer_loss": 0.1868,
-      "loss": 0.1885,
-      "short_answer_loss": NaN,
-      "step": 409,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1901,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.9439495297464512e-05,
-      "long_answer_loss": 0.1901,
-      "loss": 0.1907,
-      "short_answer_loss": NaN,
-      "step": 410,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.2132,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.94120928072064e-05,
-      "long_answer_loss": 0.2132,
-      "loss": 0.2018,
-      "short_answer_loss": NaN,
-      "step": 411,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.2384,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.9384642384659446e-05,
-      "long_answer_loss": 0.2384,
-      "loss": 0.1993,
-      "short_answer_loss": NaN,
-      "step": 412,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.2068,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.935714422018011e-05,
-      "long_answer_loss": 0.2068,
-      "loss": 0.1958,
-      "short_answer_loss": NaN,
-      "step": 413,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1957,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.932959850445594e-05,
-      "long_answer_loss": 0.1957,
-      "loss": 0.1832,
-      "short_answer_loss": NaN,
-      "step": 414,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1987,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.9302005428504216e-05,
-      "long_answer_loss": 0.1987,
-      "loss": 0.1965,
-      "short_answer_loss": NaN,
-      "step": 415,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1942,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.9274365183670645e-05,
-      "long_answer_loss": 0.1942,
-      "loss": 0.1973,
-      "short_answer_loss": NaN,
-      "step": 416,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1949,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.9246677961628028e-05,
-      "long_answer_loss": 0.1949,
-      "loss": 0.1808,
-      "short_answer_loss": NaN,
-      "step": 417,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1828,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.921894395437494e-05,
-      "long_answer_loss": 0.1828,
-      "loss": 0.1912,
-      "short_answer_loss": NaN,
-      "step": 418,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1984,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.9191163354234375e-05,
-      "long_answer_loss": 0.1984,
-      "loss": 0.1899,
-      "short_answer_loss": NaN,
-      "step": 419,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.2025,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.9163336353852447e-05,
-      "long_answer_loss": 0.2025,
-      "loss": 0.1932,
-      "short_answer_loss": NaN,
-      "step": 420,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1864,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.9135463146197018e-05,
-      "long_answer_loss": 0.1864,
-      "loss": 0.1836,
-      "short_answer_loss": NaN,
-      "step": 421,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1888,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.9107543924556388e-05,
-      "long_answer_loss": 0.1888,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 422,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1497,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.9079578882537935e-05,
-      "long_answer_loss": 0.1497,
-      "loss": 0.1897,
-      "short_answer_loss": NaN,
-      "step": 423,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1934,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.905156821406679e-05,
-      "long_answer_loss": 0.1934,
-      "loss": 0.187,
-      "short_answer_loss": NaN,
-      "step": 424,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1715,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.902351211338448e-05,
-      "long_answer_loss": 0.1715,
-      "loss": 0.179,
-      "short_answer_loss": NaN,
-      "step": 425,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1962,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.8995410775047573e-05,
-      "long_answer_loss": 0.1962,
-      "loss": 0.1918,
-      "short_answer_loss": NaN,
-      "step": 426,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.183,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.8967264393926355e-05,
-      "long_answer_loss": 0.183,
-      "loss": 0.1884,
-      "short_answer_loss": NaN,
-      "step": 427,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.2169,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.8939073165203462e-05,
-      "long_answer_loss": 0.2169,
-      "loss": 0.1868,
-      "short_answer_loss": NaN,
-      "step": 428,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1911,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.8910837284372528e-05,
-      "long_answer_loss": 0.1911,
-      "loss": 0.1747,
-      "short_answer_loss": NaN,
-      "step": 429,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1722,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.8882556947236817e-05,
-      "long_answer_loss": 0.1722,
-      "loss": 0.1815,
-      "short_answer_loss": NaN,
-      "step": 430,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1825,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.8854232349907885e-05,
-      "long_answer_loss": 0.1825,
-      "loss": 0.1875,
-      "short_answer_loss": NaN,
-      "step": 431,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1976,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.882586368880423e-05,
-      "long_answer_loss": 0.1976,
-      "loss": 0.1922,
-      "short_answer_loss": NaN,
-      "step": 432,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1526,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.8797451160649875e-05,
-      "long_answer_loss": 0.1526,
-      "loss": 0.188,
-      "short_answer_loss": NaN,
-      "step": 433,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1688,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.8768994962473085e-05,
-      "long_answer_loss": 0.1688,
-      "loss": 0.1917,
-      "short_answer_loss": NaN,
-      "step": 434,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.2068,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.8740495291604927e-05,
-      "long_answer_loss": 0.2068,
-      "loss": 0.1988,
-      "short_answer_loss": NaN,
-      "step": 435,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1973,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.8711952345677936e-05,
-      "long_answer_loss": 0.1973,
-      "loss": 0.1863,
-      "short_answer_loss": NaN,
-      "step": 436,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.2113,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.868336632262475e-05,
-      "long_answer_loss": 0.2113,
-      "loss": 0.1987,
-      "short_answer_loss": NaN,
-      "step": 437,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1502,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.8654737420676722e-05,
-      "long_answer_loss": 0.1502,
-      "loss": 0.1935,
-      "short_answer_loss": NaN,
-      "step": 438,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1857,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.8626065838362554e-05,
-      "long_answer_loss": 0.1857,
-      "loss": 0.1895,
-      "short_answer_loss": NaN,
-      "step": 439,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.2016,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.8597351774506912e-05,
-      "long_answer_loss": 0.2016,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 440,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.2299,
-      "grad_norm": 1.625,
-      "learning_rate": 1.8568595428229057e-05,
-      "long_answer_loss": 0.2299,
-      "loss": 0.1816,
-      "short_answer_loss": NaN,
-      "step": 441,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.2218,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.8539796998941456e-05,
-      "long_answer_loss": 0.2218,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 442,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.2028,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.851095668634841e-05,
-      "long_answer_loss": 0.2028,
-      "loss": 0.1904,
-      "short_answer_loss": NaN,
-      "step": 443,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.2133,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.8482074690444652e-05,
-      "long_answer_loss": 0.2133,
-      "loss": 0.1965,
-      "short_answer_loss": NaN,
-      "step": 444,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.2011,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.8453151211513984e-05,
-      "long_answer_loss": 0.2011,
-      "loss": 0.1925,
-      "short_answer_loss": NaN,
-      "step": 445,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.1733,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.8424186450127857e-05,
-      "long_answer_loss": 0.1733,
-      "loss": 0.1821,
-      "short_answer_loss": NaN,
-      "step": 446,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1843,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.839518060714401e-05,
-      "long_answer_loss": 0.1843,
-      "loss": 0.1801,
-      "short_answer_loss": NaN,
-      "step": 447,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1795,
-      "grad_norm": 1.875,
-      "learning_rate": 1.8366133883705063e-05,
-      "long_answer_loss": 0.1795,
-      "loss": 0.1842,
-      "short_answer_loss": NaN,
-      "step": 448,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.228,
-      "grad_norm": 1.9921875,
-      "learning_rate": 1.833704648123712e-05,
-      "long_answer_loss": 0.228,
-      "loss": 0.1947,
-      "short_answer_loss": NaN,
-      "step": 449,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.2036,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.830791860144838e-05,
-      "long_answer_loss": 0.2036,
-      "loss": 0.189,
-      "short_answer_loss": NaN,
-      "step": 450,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1889,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.827875044632773e-05,
-      "long_answer_loss": 0.1889,
-      "loss": 0.1839,
-      "short_answer_loss": NaN,
-      "step": 451,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.2054,
-      "grad_norm": 2.0,
-      "learning_rate": 1.824954221814335e-05,
-      "long_answer_loss": 0.2054,
-      "loss": 0.1839,
-      "short_answer_loss": NaN,
-      "step": 452,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.2043,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.822029411944131e-05,
-      "long_answer_loss": 0.2043,
-      "loss": 0.1838,
-      "short_answer_loss": NaN,
-      "step": 453,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1935,
-      "grad_norm": 1.9765625,
-      "learning_rate": 1.8191006353044165e-05,
-      "long_answer_loss": 0.1935,
-      "loss": 0.2006,
-      "short_answer_loss": NaN,
-      "step": 454,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1797,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.8161679122049545e-05,
-      "long_answer_loss": 0.1797,
-      "loss": 0.1889,
-      "short_answer_loss": NaN,
-      "step": 455,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1899,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.8132312629828758e-05,
-      "long_answer_loss": 0.1899,
-      "loss": 0.2,
-      "short_answer_loss": NaN,
-      "step": 456,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1709,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.810290708002535e-05,
-      "long_answer_loss": 0.1709,
-      "loss": 0.1758,
-      "short_answer_loss": NaN,
-      "step": 457,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1712,
-      "grad_norm": 1.625,
-      "learning_rate": 1.807346267655374e-05,
-      "long_answer_loss": 0.1712,
-      "loss": 0.1902,
-      "short_answer_loss": NaN,
-      "step": 458,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1963,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.8043979623597766e-05,
-      "long_answer_loss": 0.1963,
-      "loss": 0.1862,
-      "short_answer_loss": NaN,
-      "step": 459,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.162,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.801445812560928e-05,
-      "long_answer_loss": 0.162,
-      "loss": 0.182,
-      "short_answer_loss": NaN,
-      "step": 460,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1656,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.7984898387306743e-05,
-      "long_answer_loss": 0.1656,
-      "loss": 0.1773,
-      "short_answer_loss": NaN,
-      "step": 461,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1905,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.795530061367379e-05,
-      "long_answer_loss": 0.1905,
-      "loss": 0.1913,
-      "short_answer_loss": NaN,
-      "step": 462,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1724,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.7925665009957815e-05,
-      "long_answer_loss": 0.1724,
-      "loss": 0.1764,
-      "short_answer_loss": NaN,
-      "step": 463,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1785,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.7895991781668544e-05,
-      "long_answer_loss": 0.1785,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 464,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1875,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.786628113457662e-05,
-      "long_answer_loss": 0.1875,
-      "loss": 0.1954,
-      "short_answer_loss": NaN,
-      "step": 465,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1853,
-      "grad_norm": 1.9921875,
-      "learning_rate": 1.783653327471216e-05,
-      "long_answer_loss": 0.1853,
-      "loss": 0.1987,
-      "short_answer_loss": NaN,
-      "step": 466,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1805,
-      "grad_norm": 1.8828125,
-      "learning_rate": 1.7806748408363337e-05,
-      "long_answer_loss": 0.1805,
-      "loss": 0.1925,
-      "short_answer_loss": NaN,
-      "step": 467,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1832,
-      "grad_norm": 1.9140625,
-      "learning_rate": 1.7776926742074957e-05,
-      "long_answer_loss": 0.1832,
-      "loss": 0.1964,
-      "short_answer_loss": NaN,
-      "step": 468,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1755,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.7747068482647004e-05,
-      "long_answer_loss": 0.1755,
-      "loss": 0.1927,
-      "short_answer_loss": NaN,
-      "step": 469,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.175,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.7717173837133233e-05,
-      "long_answer_loss": 0.175,
-      "loss": 0.2062,
-      "short_answer_loss": NaN,
-      "step": 470,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.2087,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.7687243012839698e-05,
-      "long_answer_loss": 0.2087,
-      "loss": 0.2025,
-      "short_answer_loss": NaN,
-      "step": 471,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1843,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.7657276217323364e-05,
-      "long_answer_loss": 0.1843,
-      "loss": 0.1831,
-      "short_answer_loss": NaN,
-      "step": 472,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1978,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.7627273658390625e-05,
-      "long_answer_loss": 0.1978,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 473,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.201,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.7597235544095873e-05,
-      "long_answer_loss": 0.201,
-      "loss": 0.1844,
-      "short_answer_loss": NaN,
-      "step": 474,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.223,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.7567162082740082e-05,
-      "long_answer_loss": 0.223,
-      "loss": 0.1894,
-      "short_answer_loss": NaN,
-      "step": 475,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.2016,
-      "grad_norm": 1.8515625,
-      "learning_rate": 1.7537053482869316e-05,
-      "long_answer_loss": 0.2016,
-      "loss": 0.1947,
-      "short_answer_loss": NaN,
-      "step": 476,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1819,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.7506909953273327e-05,
-      "long_answer_loss": 0.1819,
-      "loss": 0.1917,
-      "short_answer_loss": NaN,
-      "step": 477,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.2318,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.7476731702984084e-05,
-      "long_answer_loss": 0.2318,
-      "loss": 0.199,
-      "short_answer_loss": NaN,
-      "step": 478,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1817,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.744651894127433e-05,
-      "long_answer_loss": 0.1817,
-      "loss": 0.1928,
-      "short_answer_loss": NaN,
-      "step": 479,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1705,
-      "grad_norm": 1.859375,
-      "learning_rate": 1.741627187765612e-05,
-      "long_answer_loss": 0.1705,
-      "loss": 0.1837,
-      "short_answer_loss": NaN,
-      "step": 480,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1718,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.7385990721879393e-05,
-      "long_answer_loss": 0.1718,
-      "loss": 0.1797,
-      "short_answer_loss": NaN,
-      "step": 481,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1963,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.735567568393049e-05,
-      "long_answer_loss": 0.1963,
-      "loss": 0.1869,
-      "short_answer_loss": NaN,
-      "step": 482,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1836,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.7325326974030717e-05,
-      "long_answer_loss": 0.1836,
-      "loss": 0.18,
-      "short_answer_loss": NaN,
-      "step": 483,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1609,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.729494480263487e-05,
-      "long_answer_loss": 0.1609,
-      "loss": 0.1895,
-      "short_answer_loss": NaN,
-      "step": 484,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1826,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.7264529380429805e-05,
-      "long_answer_loss": 0.1826,
-      "loss": 0.1836,
-      "short_answer_loss": NaN,
-      "step": 485,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.2017,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.723408091833294e-05,
-      "long_answer_loss": 0.2017,
-      "loss": 0.1864,
-      "short_answer_loss": NaN,
-      "step": 486,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1734,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.7203599627490803e-05,
-      "long_answer_loss": 0.1734,
-      "loss": 0.1876,
-      "short_answer_loss": NaN,
-      "step": 487,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1711,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.71730857192776e-05,
-      "long_answer_loss": 0.1711,
-      "loss": 0.1834,
-      "short_answer_loss": NaN,
-      "step": 488,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1615,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.7142539405293706e-05,
-      "long_answer_loss": 0.1615,
-      "loss": 0.1786,
-      "short_answer_loss": NaN,
-      "step": 489,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1881,
-      "grad_norm": 1.75,
-      "learning_rate": 1.7111960897364226e-05,
-      "long_answer_loss": 0.1881,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 490,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1699,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.7081350407537497e-05,
-      "long_answer_loss": 0.1699,
-      "loss": 0.1849,
-      "short_answer_loss": NaN,
-      "step": 491,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1865,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.7050708148083664e-05,
-      "long_answer_loss": 0.1865,
-      "loss": 0.1818,
-      "short_answer_loss": NaN,
-      "step": 492,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1678,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.7020034331493157e-05,
-      "long_answer_loss": 0.1678,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 493,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1888,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.698932917047525e-05,
-      "long_answer_loss": 0.1888,
-      "loss": 0.1779,
-      "short_answer_loss": NaN,
-      "step": 494,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.2053,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.6958592877956574e-05,
-      "long_answer_loss": 0.2053,
-      "loss": 0.1902,
-      "short_answer_loss": NaN,
-      "step": 495,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1983,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.692782566707965e-05,
-      "long_answer_loss": 0.1983,
-      "loss": 0.195,
-      "short_answer_loss": NaN,
-      "step": 496,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.2063,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.68970277512014e-05,
-      "long_answer_loss": 0.2063,
-      "loss": 0.1885,
-      "short_answer_loss": NaN,
-      "step": 497,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1664,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.6866199343891665e-05,
-      "long_answer_loss": 0.1664,
-      "loss": 0.1824,
-      "short_answer_loss": NaN,
-      "step": 498,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1684,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.6835340658931735e-05,
-      "long_answer_loss": 0.1684,
-      "loss": 0.1917,
-      "short_answer_loss": NaN,
-      "step": 499,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1759,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.680445191031287e-05,
-      "long_answer_loss": 0.1759,
-      "loss": 0.1744,
-      "short_answer_loss": NaN,
-      "step": 500,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1767,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.6773533312234794e-05,
-      "long_answer_loss": 0.1767,
-      "loss": 0.1707,
-      "short_answer_loss": NaN,
-      "step": 501,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.2144,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.6742585079104234e-05,
-      "long_answer_loss": 0.2144,
-      "loss": 0.1946,
-      "short_answer_loss": NaN,
-      "step": 502,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1822,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.6711607425533422e-05,
-      "long_answer_loss": 0.1822,
-      "loss": 0.1989,
-      "short_answer_loss": NaN,
-      "step": 503,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1467,
-      "grad_norm": 1.75,
-      "learning_rate": 1.66806005663386e-05,
-      "long_answer_loss": 0.1467,
-      "loss": 0.1771,
-      "short_answer_loss": NaN,
-      "step": 504,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1862,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.664956471653855e-05,
-      "long_answer_loss": 0.1862,
-      "loss": 0.1844,
-      "short_answer_loss": NaN,
-      "step": 505,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1797,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.661850009135308e-05,
-      "long_answer_loss": 0.1797,
-      "loss": 0.1909,
-      "short_answer_loss": NaN,
-      "step": 506,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1752,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.6587406906201546e-05,
-      "long_answer_loss": 0.1752,
-      "loss": 0.1763,
-      "short_answer_loss": NaN,
-      "step": 507,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1691,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.6556285376701357e-05,
-      "long_answer_loss": 0.1691,
-      "loss": 0.1866,
-      "short_answer_loss": NaN,
-      "step": 508,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1908,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.652513571866648e-05,
-      "long_answer_loss": 0.1908,
-      "loss": 0.1844,
-      "short_answer_loss": NaN,
-      "step": 509,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.2111,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.6493958148105934e-05,
-      "long_answer_loss": 0.2111,
-      "loss": 0.1786,
-      "short_answer_loss": NaN,
-      "step": 510,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1876,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.64627528812223e-05,
-      "long_answer_loss": 0.1876,
-      "loss": 0.1928,
-      "short_answer_loss": NaN,
-      "step": 511,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1758,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.6431520134410222e-05,
-      "long_answer_loss": 0.1758,
-      "loss": 0.1751,
-      "short_answer_loss": NaN,
-      "step": 512,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1475,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.6400260124254905e-05,
-      "long_answer_loss": 0.1475,
-      "loss": 0.1894,
-      "short_answer_loss": NaN,
-      "step": 513,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1734,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.6368973067530618e-05,
-      "long_answer_loss": 0.1734,
-      "loss": 0.1699,
-      "short_answer_loss": NaN,
-      "step": 514,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.2011,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.6337659181199184e-05,
-      "long_answer_loss": 0.2011,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 515,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.2202,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.6306318682408468e-05,
-      "long_answer_loss": 0.2202,
-      "loss": 0.1879,
-      "short_answer_loss": NaN,
-      "step": 516,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1634,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.6274951788490887e-05,
-      "long_answer_loss": 0.1634,
-      "loss": 0.1849,
-      "short_answer_loss": NaN,
-      "step": 517,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.2103,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.6243558716961904e-05,
-      "long_answer_loss": 0.2103,
-      "loss": 0.1863,
-      "short_answer_loss": NaN,
-      "step": 518,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1896,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.6212139685518503e-05,
-      "long_answer_loss": 0.1896,
-      "loss": 0.1803,
-      "short_answer_loss": NaN,
-      "step": 519,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.204,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.6180694912037687e-05,
-      "long_answer_loss": 0.204,
-      "loss": 0.2032,
-      "short_answer_loss": NaN,
-      "step": 520,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1782,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.6149224614574974e-05,
-      "long_answer_loss": 0.1782,
-      "loss": 0.1894,
-      "short_answer_loss": NaN,
-      "step": 521,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1584,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.6117729011362882e-05,
-      "long_answer_loss": 0.1584,
-      "loss": 0.1796,
-      "short_answer_loss": NaN,
-      "step": 522,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1931,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.6086208320809403e-05,
-      "long_answer_loss": 0.1931,
-      "loss": 0.1804,
-      "short_answer_loss": NaN,
-      "step": 523,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.174,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.6054662761496504e-05,
-      "long_answer_loss": 0.174,
-      "loss": 0.1843,
-      "short_answer_loss": NaN,
-      "step": 524,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1966,
-      "grad_norm": 1.625,
-      "learning_rate": 1.6023092552178607e-05,
-      "long_answer_loss": 0.1966,
-      "loss": 0.1843,
-      "short_answer_loss": NaN,
-      "step": 525,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.201,
-      "grad_norm": 1.625,
-      "learning_rate": 1.5991497911781073e-05,
-      "long_answer_loss": 0.201,
-      "loss": 0.177,
-      "short_answer_loss": NaN,
-      "step": 526,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1943,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.595987905939868e-05,
-      "long_answer_loss": 0.1943,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 527,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1929,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.5928236214294097e-05,
-      "long_answer_loss": 0.1929,
-      "loss": 0.1858,
-      "short_answer_loss": NaN,
-      "step": 528,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1678,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.5896569595896384e-05,
-      "long_answer_loss": 0.1678,
-      "loss": 0.1783,
-      "short_answer_loss": NaN,
-      "step": 529,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1622,
-      "grad_norm": 1.8671875,
-      "learning_rate": 1.5864879423799454e-05,
-      "long_answer_loss": 0.1622,
-      "loss": 0.1778,
-      "short_answer_loss": NaN,
-      "step": 530,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1978,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.5833165917760544e-05,
-      "long_answer_loss": 0.1978,
-      "loss": 0.1749,
-      "short_answer_loss": NaN,
-      "step": 531,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1788,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.5801429297698725e-05,
-      "long_answer_loss": 0.1788,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 532,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.195,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.576966978369333e-05,
-      "long_answer_loss": 0.195,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 533,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1846,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.5737887595982466e-05,
-      "long_answer_loss": 0.1846,
-      "loss": 0.1779,
-      "short_answer_loss": NaN,
-      "step": 534,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1965,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.570608295496146e-05,
-      "long_answer_loss": 0.1965,
-      "loss": 0.1819,
-      "short_answer_loss": NaN,
-      "step": 535,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.2053,
-      "grad_norm": 1.625,
-      "learning_rate": 1.567425608118136e-05,
-      "long_answer_loss": 0.2053,
-      "loss": 0.1809,
-      "short_answer_loss": NaN,
-      "step": 536,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.2033,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.5642407195347367e-05,
-      "long_answer_loss": 0.2033,
-      "loss": 0.1914,
-      "short_answer_loss": NaN,
-      "step": 537,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.181,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.561053651831734e-05,
-      "long_answer_loss": 0.181,
-      "loss": 0.1728,
-      "short_answer_loss": NaN,
-      "step": 538,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1719,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.557864427110025e-05,
-      "long_answer_loss": 0.1719,
-      "loss": 0.2032,
-      "short_answer_loss": NaN,
-      "step": 539,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1808,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.5546730674854643e-05,
-      "long_answer_loss": 0.1808,
-      "loss": 0.1807,
-      "short_answer_loss": NaN,
-      "step": 540,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.189,
-      "grad_norm": 1.75,
-      "learning_rate": 1.5514795950887107e-05,
-      "long_answer_loss": 0.189,
-      "loss": 0.1861,
-      "short_answer_loss": NaN,
-      "step": 541,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.2012,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.5482840320650756e-05,
-      "long_answer_loss": 0.2012,
-      "loss": 0.1855,
-      "short_answer_loss": NaN,
-      "step": 542,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1631,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.545086400574367e-05,
-      "long_answer_loss": 0.1631,
-      "loss": 0.1741,
-      "short_answer_loss": NaN,
-      "step": 543,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1732,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.541886722790736e-05,
-      "long_answer_loss": 0.1732,
-      "loss": 0.1891,
-      "short_answer_loss": NaN,
-      "step": 544,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.2084,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.538685020902525e-05,
-      "long_answer_loss": 0.2084,
-      "loss": 0.1823,
-      "short_answer_loss": NaN,
-      "step": 545,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1735,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.5354813171121133e-05,
-      "long_answer_loss": 0.1735,
-      "loss": 0.1918,
-      "short_answer_loss": NaN,
-      "step": 546,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1942,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.53227563363576e-05,
-      "long_answer_loss": 0.1942,
-      "loss": 0.1803,
-      "short_answer_loss": NaN,
-      "step": 547,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1862,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.529067992703455e-05,
-      "long_answer_loss": 0.1862,
-      "loss": 0.1843,
-      "short_answer_loss": NaN,
-      "step": 548,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1778,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.5258584165587614e-05,
-      "long_answer_loss": 0.1778,
-      "loss": 0.1847,
-      "short_answer_loss": NaN,
-      "step": 549,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1799,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.5226469274586617e-05,
-      "long_answer_loss": 0.1799,
-      "loss": 0.1859,
-      "short_answer_loss": NaN,
-      "step": 550,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1702,
-      "grad_norm": 1.75,
-      "learning_rate": 1.5194335476734036e-05,
-      "long_answer_loss": 0.1702,
-      "loss": 0.1678,
-      "short_answer_loss": NaN,
-      "step": 551,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1955,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.5162182994863474e-05,
-      "long_answer_loss": 0.1955,
-      "loss": 0.1836,
-      "short_answer_loss": NaN,
-      "step": 552,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.2061,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.5130012051938089e-05,
-      "long_answer_loss": 0.2061,
-      "loss": 0.1799,
-      "short_answer_loss": NaN,
-      "step": 553,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1503,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.5097822871049055e-05,
-      "long_answer_loss": 0.1503,
-      "loss": 0.1681,
-      "short_answer_loss": NaN,
-      "step": 554,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1901,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.506561567541402e-05,
-      "long_answer_loss": 0.1901,
-      "loss": 0.1692,
-      "short_answer_loss": NaN,
-      "step": 555,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1438,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.503339068837557e-05,
-      "long_answer_loss": 0.1438,
-      "loss": 0.1724,
-      "short_answer_loss": NaN,
-      "step": 556,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1826,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.500114813339965e-05,
-      "long_answer_loss": 0.1826,
-      "loss": 0.1844,
-      "short_answer_loss": NaN,
-      "step": 557,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1934,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.4968888234074027e-05,
-      "long_answer_loss": 0.1934,
-      "loss": 0.1834,
-      "short_answer_loss": NaN,
-      "step": 558,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1649,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.4936611214106763e-05,
-      "long_answer_loss": 0.1649,
-      "loss": 0.1748,
-      "short_answer_loss": NaN,
-      "step": 559,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.203,
-      "grad_norm": 1.75,
-      "learning_rate": 1.4904317297324633e-05,
-      "long_answer_loss": 0.203,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 560,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1788,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.487200670767158e-05,
-      "long_answer_loss": 0.1788,
-      "loss": 0.1727,
-      "short_answer_loss": NaN,
-      "step": 561,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1531,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.4839679669207168e-05,
-      "long_answer_loss": 0.1531,
-      "loss": 0.1823,
-      "short_answer_loss": NaN,
-      "step": 562,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.196,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.4807336406105032e-05,
-      "long_answer_loss": 0.196,
-      "loss": 0.168,
-      "short_answer_loss": NaN,
-      "step": 563,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1461,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.4774977142651316e-05,
-      "long_answer_loss": 0.1461,
-      "loss": 0.1678,
-      "short_answer_loss": NaN,
-      "step": 564,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1903,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.4742602103243113e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1829,
-      "short_answer_loss": NaN,
-      "step": 565,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.195,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.4710211512386927e-05,
-      "long_answer_loss": 0.195,
-      "loss": 0.1775,
-      "short_answer_loss": NaN,
-      "step": 566,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1732,
-      "grad_norm": 1.8671875,
-      "learning_rate": 1.467780559469709e-05,
-      "long_answer_loss": 0.1732,
-      "loss": 0.1705,
-      "short_answer_loss": NaN,
-      "step": 567,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.181,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.4645384574894224e-05,
-      "long_answer_loss": 0.181,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 568,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1779,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.4612948677803687e-05,
-      "long_answer_loss": 0.1779,
-      "loss": 0.1809,
-      "short_answer_loss": NaN,
-      "step": 569,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1518,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.4580498128353992e-05,
-      "long_answer_loss": 0.1518,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 570,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1903,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.4548033151575264e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1726,
-      "short_answer_loss": NaN,
-      "step": 571,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.2106,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.4515553972597672e-05,
-      "long_answer_loss": 0.2106,
-      "loss": 0.1886,
-      "short_answer_loss": NaN,
-      "step": 572,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1501,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.4483060816649885e-05,
-      "long_answer_loss": 0.1501,
-      "loss": 0.1799,
-      "short_answer_loss": NaN,
-      "step": 573,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1709,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.4450553909057473e-05,
-      "long_answer_loss": 0.1709,
-      "loss": 0.1709,
-      "short_answer_loss": NaN,
-      "step": 574,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.2119,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.4418033475241388e-05,
-      "long_answer_loss": 0.2119,
-      "loss": 0.1827,
-      "short_answer_loss": NaN,
-      "step": 575,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1803,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.4385499740716369e-05,
-      "long_answer_loss": 0.1803,
-      "loss": 0.1831,
-      "short_answer_loss": NaN,
-      "step": 576,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1662,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.4352952931089392e-05,
-      "long_answer_loss": 0.1662,
-      "loss": 0.1784,
-      "short_answer_loss": NaN,
-      "step": 577,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.2044,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.4320393272058103e-05,
-      "long_answer_loss": 0.2044,
-      "loss": 0.1848,
-      "short_answer_loss": NaN,
-      "step": 578,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1684,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.4287820989409257e-05,
-      "long_answer_loss": 0.1684,
-      "loss": 0.1852,
-      "short_answer_loss": NaN,
-      "step": 579,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1639,
-      "grad_norm": 1.625,
-      "learning_rate": 1.4255236309017145e-05,
-      "long_answer_loss": 0.1639,
-      "loss": 0.1847,
-      "short_answer_loss": NaN,
-      "step": 580,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1733,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.4222639456842024e-05,
-      "long_answer_loss": 0.1733,
-      "loss": 0.1781,
-      "short_answer_loss": NaN,
-      "step": 581,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1701,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.4190030658928568e-05,
-      "long_answer_loss": 0.1701,
-      "loss": 0.1802,
-      "short_answer_loss": NaN,
-      "step": 582,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.175,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.415741014140429e-05,
-      "long_answer_loss": 0.175,
-      "loss": 0.182,
-      "short_answer_loss": NaN,
-      "step": 583,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1588,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.4124778130477962e-05,
-      "long_answer_loss": 0.1588,
-      "loss": 0.1865,
-      "short_answer_loss": NaN,
-      "step": 584,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1698,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.4092134852438054e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.1725,
-      "short_answer_loss": NaN,
-      "step": 585,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.2202,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.4059480533651201e-05,
-      "long_answer_loss": 0.2202,
-      "loss": 0.1891,
-      "short_answer_loss": NaN,
-      "step": 586,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1516,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.4026815400560561e-05,
-      "long_answer_loss": 0.1516,
-      "loss": 0.1634,
-      "short_answer_loss": NaN,
-      "step": 587,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1903,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.3994139679684306e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1833,
-      "short_answer_loss": NaN,
-      "step": 588,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1829,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.3961453597614025e-05,
-      "long_answer_loss": 0.1829,
-      "loss": 0.1844,
-      "short_answer_loss": NaN,
-      "step": 589,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1826,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.3928757381013158e-05,
-      "long_answer_loss": 0.1826,
-      "loss": 0.179,
-      "short_answer_loss": NaN,
-      "step": 590,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1558,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.3896051256615423e-05,
-      "long_answer_loss": 0.1558,
-      "loss": 0.1763,
-      "short_answer_loss": NaN,
-      "step": 591,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.173,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.3863335451223235e-05,
-      "long_answer_loss": 0.173,
-      "loss": 0.1724,
-      "short_answer_loss": NaN,
-      "step": 592,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.2119,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.3830610191706162e-05,
-      "long_answer_loss": 0.2119,
-      "loss": 0.178,
-      "short_answer_loss": NaN,
-      "step": 593,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1676,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.3797875704999319e-05,
-      "long_answer_loss": 0.1676,
-      "loss": 0.1809,
-      "short_answer_loss": NaN,
-      "step": 594,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1737,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.3765132218101806e-05,
-      "long_answer_loss": 0.1737,
-      "loss": 0.1718,
-      "short_answer_loss": NaN,
-      "step": 595,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1874,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.3732379958075142e-05,
-      "long_answer_loss": 0.1874,
-      "loss": 0.1743,
-      "short_answer_loss": NaN,
-      "step": 596,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1611,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.3699619152041681e-05,
-      "long_answer_loss": 0.1611,
-      "loss": 0.1879,
-      "short_answer_loss": NaN,
-      "step": 597,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.173,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.366685002718304e-05,
-      "long_answer_loss": 0.173,
-      "loss": 0.1777,
-      "short_answer_loss": NaN,
-      "step": 598,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1729,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.3634072810738518e-05,
-      "long_answer_loss": 0.1729,
-      "loss": 0.1718,
-      "short_answer_loss": NaN,
-      "step": 599,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1439,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.3601287730003534e-05,
-      "long_answer_loss": 0.1439,
-      "loss": 0.1749,
-      "short_answer_loss": NaN,
-      "step": 600,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1837,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.3568495012328041e-05,
-      "long_answer_loss": 0.1837,
-      "loss": 0.182,
-      "short_answer_loss": NaN,
-      "step": 601,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1871,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.3535694885114947e-05,
-      "long_answer_loss": 0.1871,
-      "loss": 0.1758,
-      "short_answer_loss": NaN,
-      "step": 602,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1669,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.350288757581854e-05,
-      "long_answer_loss": 0.1669,
-      "loss": 0.1647,
-      "short_answer_loss": NaN,
-      "step": 603,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1866,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.3470073311942918e-05,
-      "long_answer_loss": 0.1866,
-      "loss": 0.1774,
-      "short_answer_loss": NaN,
-      "step": 604,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1827,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.3437252321040406e-05,
-      "long_answer_loss": 0.1827,
-      "loss": 0.1828,
-      "short_answer_loss": NaN,
-      "step": 605,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.197,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.3404424830709978e-05,
-      "long_answer_loss": 0.197,
-      "loss": 0.1771,
-      "short_answer_loss": NaN,
-      "step": 606,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1834,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.337159106859567e-05,
-      "long_answer_loss": 0.1834,
-      "loss": 0.1801,
-      "short_answer_loss": NaN,
-      "step": 607,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1698,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.333875126238502e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.18,
-      "short_answer_loss": NaN,
-      "step": 608,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1541,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.330590563980747e-05,
-      "long_answer_loss": 0.1541,
-      "loss": 0.1847,
-      "short_answer_loss": NaN,
-      "step": 609,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1769,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.3273054428632814e-05,
-      "long_answer_loss": 0.1769,
-      "loss": 0.1792,
-      "short_answer_loss": NaN,
-      "step": 610,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.2038,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.3240197856669576e-05,
-      "long_answer_loss": 0.2038,
-      "loss": 0.1751,
-      "short_answer_loss": NaN,
-      "step": 611,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1696,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.3207336151763469e-05,
-      "long_answer_loss": 0.1696,
-      "loss": 0.1659,
-      "short_answer_loss": NaN,
-      "step": 612,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1555,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.3174469541795806e-05,
-      "long_answer_loss": 0.1555,
-      "loss": 0.1661,
-      "short_answer_loss": NaN,
-      "step": 613,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1672,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.3141598254681903e-05,
-      "long_answer_loss": 0.1672,
-      "loss": 0.1765,
-      "short_answer_loss": NaN,
-      "step": 614,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1691,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.3108722518369507e-05,
-      "long_answer_loss": 0.1691,
-      "loss": 0.1751,
-      "short_answer_loss": NaN,
-      "step": 615,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1399,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.3075842560837229e-05,
-      "long_answer_loss": 0.1399,
-      "loss": 0.1428,
-      "short_answer_loss": NaN,
-      "step": 616,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.0948,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.3042958610092953e-05,
-      "long_answer_loss": 0.0948,
-      "loss": 0.0965,
-      "short_answer_loss": NaN,
-      "step": 617,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1109,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.3010070894172247e-05,
-      "long_answer_loss": 0.1109,
-      "loss": 0.1022,
-      "short_answer_loss": NaN,
-      "step": 618,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0929,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.297717964113678e-05,
-      "long_answer_loss": 0.0929,
-      "loss": 0.0954,
-      "short_answer_loss": NaN,
-      "step": 619,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0971,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.294428507907278e-05,
-      "long_answer_loss": 0.0971,
-      "loss": 0.1051,
-      "short_answer_loss": NaN,
-      "step": 620,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0911,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.291138743608939e-05,
-      "long_answer_loss": 0.0911,
-      "loss": 0.0948,
-      "short_answer_loss": NaN,
-      "step": 621,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0939,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.2878486940317128e-05,
-      "long_answer_loss": 0.0939,
-      "loss": 0.0926,
-      "short_answer_loss": NaN,
-      "step": 622,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0889,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.2845583819906303e-05,
-      "long_answer_loss": 0.0889,
-      "loss": 0.0924,
-      "short_answer_loss": NaN,
-      "step": 623,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.1197,
-      "grad_norm": 1.75,
-      "learning_rate": 1.2812678303025419e-05,
-      "long_answer_loss": 0.1197,
-      "loss": 0.1079,
-      "short_answer_loss": NaN,
-      "step": 624,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.105,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.2779770617859593e-05,
-      "long_answer_loss": 0.105,
-      "loss": 0.0952,
-      "short_answer_loss": NaN,
-      "step": 625,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.1089,
-      "grad_norm": 1.625,
-      "learning_rate": 1.2746860992608987e-05,
-      "long_answer_loss": 0.1089,
-      "loss": 0.0861,
-      "short_answer_loss": NaN,
-      "step": 626,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.1044,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.2713949655487206e-05,
-      "long_answer_loss": 0.1044,
-      "loss": 0.1001,
-      "short_answer_loss": NaN,
-      "step": 627,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.1123,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.2681036834719742e-05,
-      "long_answer_loss": 0.1123,
-      "loss": 0.1015,
-      "short_answer_loss": NaN,
-      "step": 628,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.1049,
-      "grad_norm": 1.625,
-      "learning_rate": 1.2648122758542358e-05,
-      "long_answer_loss": 0.1049,
-      "loss": 0.0966,
-      "short_answer_loss": NaN,
-      "step": 629,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0878,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.2615207655199534e-05,
-      "long_answer_loss": 0.0878,
-      "loss": 0.0913,
-      "short_answer_loss": NaN,
-      "step": 630,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0944,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.2582291752942876e-05,
-      "long_answer_loss": 0.0944,
-      "loss": 0.096,
-      "short_answer_loss": NaN,
-      "step": 631,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0822,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.2549375280029513e-05,
-      "long_answer_loss": 0.0822,
-      "loss": 0.0907,
-      "short_answer_loss": NaN,
-      "step": 632,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.1061,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.2516458464720552e-05,
-      "long_answer_loss": 0.1061,
-      "loss": 0.0994,
-      "short_answer_loss": NaN,
-      "step": 633,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0888,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.2483541535279453e-05,
-      "long_answer_loss": 0.0888,
-      "loss": 0.0921,
-      "short_answer_loss": NaN,
-      "step": 634,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0851,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.2450624719970487e-05,
-      "long_answer_loss": 0.0851,
-      "loss": 0.0929,
-      "short_answer_loss": NaN,
-      "step": 635,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.1028,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.2417708247057127e-05,
-      "long_answer_loss": 0.1028,
-      "loss": 0.0945,
-      "short_answer_loss": NaN,
-      "step": 636,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.1081,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.2384792344800467e-05,
-      "long_answer_loss": 0.1081,
-      "loss": 0.0908,
-      "short_answer_loss": NaN,
-      "step": 637,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.1061,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.2351877241457644e-05,
-      "long_answer_loss": 0.1061,
-      "loss": 0.0998,
-      "short_answer_loss": NaN,
-      "step": 638,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0892,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.2318963165280264e-05,
-      "long_answer_loss": 0.0892,
-      "loss": 0.0832,
-      "short_answer_loss": NaN,
-      "step": 639,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0781,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.22860503445128e-05,
-      "long_answer_loss": 0.0781,
-      "loss": 0.0913,
-      "short_answer_loss": NaN,
-      "step": 640,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0827,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.2253139007391018e-05,
-      "long_answer_loss": 0.0827,
-      "loss": 0.0909,
-      "short_answer_loss": NaN,
-      "step": 641,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0894,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.222022938214041e-05,
-      "long_answer_loss": 0.0894,
-      "loss": 0.0957,
-      "short_answer_loss": NaN,
-      "step": 642,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0977,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.2187321696974584e-05,
-      "long_answer_loss": 0.0977,
-      "loss": 0.0945,
-      "short_answer_loss": NaN,
-      "step": 643,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.1224,
-      "grad_norm": 1.5,
-      "learning_rate": 1.21544161800937e-05,
-      "long_answer_loss": 0.1224,
-      "loss": 0.0942,
-      "short_answer_loss": NaN,
-      "step": 644,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0931,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2121513059682873e-05,
-      "long_answer_loss": 0.0931,
-      "loss": 0.0946,
-      "short_answer_loss": NaN,
-      "step": 645,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.1158,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.2088612563910615e-05,
-      "long_answer_loss": 0.1158,
-      "loss": 0.1018,
-      "short_answer_loss": NaN,
-      "step": 646,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0816,
-      "grad_norm": 1.5,
-      "learning_rate": 1.2055714920927221e-05,
-      "long_answer_loss": 0.0816,
-      "loss": 0.0944,
-      "short_answer_loss": NaN,
-      "step": 647,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0927,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.202282035886322e-05,
-      "long_answer_loss": 0.0927,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 648,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.096,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.1989929105827757e-05,
-      "long_answer_loss": 0.096,
-      "loss": 0.0928,
-      "short_answer_loss": NaN,
-      "step": 649,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.1016,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.1957041389907051e-05,
-      "long_answer_loss": 0.1016,
-      "loss": 0.0926,
-      "short_answer_loss": NaN,
-      "step": 650,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.1134,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.1924157439162774e-05,
-      "long_answer_loss": 0.1134,
-      "loss": 0.0901,
-      "short_answer_loss": NaN,
-      "step": 651,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0934,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.1891277481630497e-05,
-      "long_answer_loss": 0.0934,
-      "loss": 0.0917,
-      "short_answer_loss": NaN,
-      "step": 652,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0942,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.1858401745318105e-05,
-      "long_answer_loss": 0.0942,
-      "loss": 0.0914,
-      "short_answer_loss": NaN,
-      "step": 653,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0835,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.1825530458204193e-05,
-      "long_answer_loss": 0.0835,
-      "loss": 0.0885,
-      "short_answer_loss": NaN,
-      "step": 654,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0952,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.1792663848236529e-05,
-      "long_answer_loss": 0.0952,
-      "loss": 0.0942,
-      "short_answer_loss": NaN,
-      "step": 655,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.1325,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.1759802143330429e-05,
-      "long_answer_loss": 0.1325,
-      "loss": 0.1,
-      "short_answer_loss": NaN,
-      "step": 656,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.086,
-      "grad_norm": 1.5,
-      "learning_rate": 1.172694557136719e-05,
-      "long_answer_loss": 0.086,
-      "loss": 0.093,
-      "short_answer_loss": NaN,
-      "step": 657,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0779,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.1694094360192532e-05,
-      "long_answer_loss": 0.0779,
-      "loss": 0.091,
-      "short_answer_loss": NaN,
-      "step": 658,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0932,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.1661248737614989e-05,
-      "long_answer_loss": 0.0932,
-      "loss": 0.0948,
-      "short_answer_loss": NaN,
-      "step": 659,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0828,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.1628408931404336e-05,
-      "long_answer_loss": 0.0828,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 660,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.1188,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.1595575169290022e-05,
-      "long_answer_loss": 0.1188,
-      "loss": 0.0959,
-      "short_answer_loss": NaN,
-      "step": 661,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0722,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.1562747678959593e-05,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0897,
-      "short_answer_loss": NaN,
-      "step": 662,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.1015,
-      "grad_norm": 1.625,
-      "learning_rate": 1.1529926688057083e-05,
-      "long_answer_loss": 0.1015,
-      "loss": 0.0949,
-      "short_answer_loss": NaN,
-      "step": 663,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.098,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.149711242418146e-05,
-      "long_answer_loss": 0.098,
-      "loss": 0.0972,
-      "short_answer_loss": NaN,
-      "step": 664,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0964,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.1464305114885057e-05,
-      "long_answer_loss": 0.0964,
-      "loss": 0.0938,
-      "short_answer_loss": NaN,
-      "step": 665,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0898,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.1431504987671963e-05,
-      "long_answer_loss": 0.0898,
-      "loss": 0.0874,
-      "short_answer_loss": NaN,
-      "step": 666,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0903,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.1398712269996468e-05,
-      "long_answer_loss": 0.0903,
-      "loss": 0.0954,
-      "short_answer_loss": NaN,
-      "step": 667,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0939,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.1365927189261482e-05,
-      "long_answer_loss": 0.0939,
-      "loss": 0.0982,
-      "short_answer_loss": NaN,
-      "step": 668,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.1123,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.1333149972816961e-05,
-      "long_answer_loss": 0.1123,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 669,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0926,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.130038084795832e-05,
-      "long_answer_loss": 0.0926,
-      "loss": 0.0901,
-      "short_answer_loss": NaN,
-      "step": 670,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0919,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.126762004192486e-05,
-      "long_answer_loss": 0.0919,
-      "loss": 0.0873,
-      "short_answer_loss": NaN,
-      "step": 671,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0889,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.1234867781898195e-05,
-      "long_answer_loss": 0.0889,
-      "loss": 0.0982,
-      "short_answer_loss": NaN,
-      "step": 672,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0893,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.1202124295000686e-05,
-      "long_answer_loss": 0.0893,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 673,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.1006,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.1169389808293836e-05,
-      "long_answer_loss": 0.1006,
-      "loss": 0.0916,
-      "short_answer_loss": NaN,
-      "step": 674,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.1058,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.1136664548776766e-05,
-      "long_answer_loss": 0.1058,
-      "loss": 0.09,
-      "short_answer_loss": NaN,
-      "step": 675,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.1017,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.1103948743384582e-05,
-      "long_answer_loss": 0.1017,
-      "loss": 0.0946,
-      "short_answer_loss": NaN,
-      "step": 676,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0869,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.1071242618986844e-05,
-      "long_answer_loss": 0.0869,
-      "loss": 0.0928,
-      "short_answer_loss": NaN,
-      "step": 677,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.083,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.1038546402385978e-05,
-      "long_answer_loss": 0.083,
-      "loss": 0.0897,
-      "short_answer_loss": NaN,
-      "step": 678,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0731,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.1005860320315697e-05,
-      "long_answer_loss": 0.0731,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 679,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0819,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.0973184599439443e-05,
-      "long_answer_loss": 0.0819,
-      "loss": 0.0897,
-      "short_answer_loss": NaN,
-      "step": 680,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.1036,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.09405194663488e-05,
-      "long_answer_loss": 0.1036,
-      "loss": 0.0921,
-      "short_answer_loss": NaN,
-      "step": 681,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0792,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.0907865147561944e-05,
-      "long_answer_loss": 0.0792,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 682,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.103,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0875221869522043e-05,
-      "long_answer_loss": 0.103,
-      "loss": 0.0947,
-      "short_answer_loss": NaN,
-      "step": 683,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0753,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.0842589858595714e-05,
-      "long_answer_loss": 0.0753,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 684,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0965,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0809969341071433e-05,
-      "long_answer_loss": 0.0965,
-      "loss": 0.088,
-      "short_answer_loss": NaN,
-      "step": 685,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0702,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.0777360543157978e-05,
-      "long_answer_loss": 0.0702,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 686,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0926,
-      "grad_norm": 1.5,
-      "learning_rate": 1.074476369098286e-05,
-      "long_answer_loss": 0.0926,
-      "loss": 0.0914,
-      "short_answer_loss": NaN,
-      "step": 687,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0942,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0712179010590742e-05,
-      "long_answer_loss": 0.0942,
-      "loss": 0.09,
-      "short_answer_loss": NaN,
-      "step": 688,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0905,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.0679606727941897e-05,
-      "long_answer_loss": 0.0905,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 689,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0831,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.0647047068910613e-05,
-      "long_answer_loss": 0.0831,
-      "loss": 0.0872,
-      "short_answer_loss": NaN,
-      "step": 690,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0856,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.0614500259283632e-05,
-      "long_answer_loss": 0.0856,
-      "loss": 0.0904,
-      "short_answer_loss": NaN,
-      "step": 691,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0705,
-      "grad_norm": 1.5,
-      "learning_rate": 1.0581966524758615e-05,
-      "long_answer_loss": 0.0705,
-      "loss": 0.0858,
-      "short_answer_loss": NaN,
-      "step": 692,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0898,
-      "grad_norm": 1.5,
-      "learning_rate": 1.054944609094253e-05,
-      "long_answer_loss": 0.0898,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 693,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0935,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.051693918335012e-05,
-      "long_answer_loss": 0.0935,
-      "loss": 0.0948,
-      "short_answer_loss": NaN,
-      "step": 694,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.069,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.0484446027402326e-05,
-      "long_answer_loss": 0.069,
-      "loss": 0.0836,
-      "short_answer_loss": NaN,
-      "step": 695,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0909,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.0451966848424737e-05,
-      "long_answer_loss": 0.0909,
-      "loss": 0.0887,
-      "short_answer_loss": NaN,
-      "step": 696,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0903,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.0419501871646012e-05,
-      "long_answer_loss": 0.0903,
-      "loss": 0.098,
-      "short_answer_loss": NaN,
-      "step": 697,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.098,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.0387051322196314e-05,
-      "long_answer_loss": 0.098,
-      "loss": 0.0919,
-      "short_answer_loss": NaN,
-      "step": 698,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.097,
-      "grad_norm": 1.625,
-      "learning_rate": 1.0354615425105779e-05,
-      "long_answer_loss": 0.097,
-      "loss": 0.0908,
-      "short_answer_loss": NaN,
-      "step": 699,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0693,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.0322194405302917e-05,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 700,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.1016,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.0289788487613074e-05,
-      "long_answer_loss": 0.1016,
-      "loss": 0.0888,
-      "short_answer_loss": NaN,
-      "step": 701,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0705,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.0257397896756884e-05,
-      "long_answer_loss": 0.0705,
-      "loss": 0.0971,
-      "short_answer_loss": NaN,
-      "step": 702,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0892,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.0225022857348685e-05,
-      "long_answer_loss": 0.0892,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 703,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.1117,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.0192663593894971e-05,
-      "long_answer_loss": 0.1117,
-      "loss": 0.0937,
-      "short_answer_loss": NaN,
-      "step": 704,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0962,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.0160320330792835e-05,
-      "long_answer_loss": 0.0962,
-      "loss": 0.0959,
-      "short_answer_loss": NaN,
-      "step": 705,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.078,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.0127993292328427e-05,
-      "long_answer_loss": 0.078,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 706,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0944,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.0095682702675375e-05,
-      "long_answer_loss": 0.0944,
-      "loss": 0.0924,
-      "short_answer_loss": NaN,
-      "step": 707,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0743,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.0063388785893236e-05,
-      "long_answer_loss": 0.0743,
-      "loss": 0.0943,
-      "short_answer_loss": NaN,
-      "step": 708,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0889,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0031111765925974e-05,
-      "long_answer_loss": 0.0889,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 709,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.1036,
-      "grad_norm": 1.515625,
-      "learning_rate": 9.998851866600353e-06,
-      "long_answer_loss": 0.1036,
-      "loss": 0.0873,
-      "short_answer_loss": NaN,
-      "step": 710,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0933,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.966609311624431e-06,
-      "long_answer_loss": 0.0933,
-      "loss": 0.092,
-      "short_answer_loss": NaN,
-      "step": 711,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0835,
-      "grad_norm": 1.4921875,
-      "learning_rate": 9.93438432458598e-06,
-      "long_answer_loss": 0.0835,
-      "loss": 0.085,
-      "short_answer_loss": NaN,
-      "step": 712,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.092,
-      "grad_norm": 1.4921875,
-      "learning_rate": 9.90217712895095e-06,
-      "long_answer_loss": 0.092,
-      "loss": 0.0888,
-      "short_answer_loss": NaN,
-      "step": 713,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0753,
-      "grad_norm": 1.5859375,
-      "learning_rate": 9.869987948061916e-06,
-      "long_answer_loss": 0.0753,
-      "loss": 0.0919,
-      "short_answer_loss": NaN,
-      "step": 714,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0875,
-      "grad_norm": 1.7109375,
-      "learning_rate": 9.837817005136525e-06,
-      "long_answer_loss": 0.0875,
-      "loss": 0.0911,
-      "short_answer_loss": NaN,
-      "step": 715,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0663,
-      "grad_norm": 1.515625,
-      "learning_rate": 9.805664523265965e-06,
-      "long_answer_loss": 0.0663,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 716,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.1043,
-      "grad_norm": 1.484375,
-      "learning_rate": 9.773530725413386e-06,
-      "long_answer_loss": 0.1043,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 717,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0888,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.741415834412389e-06,
-      "long_answer_loss": 0.0888,
-      "loss": 0.0877,
-      "short_answer_loss": NaN,
-      "step": 718,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.1257,
-      "grad_norm": 1.6328125,
-      "learning_rate": 9.709320072965451e-06,
-      "long_answer_loss": 0.1257,
-      "loss": 0.0957,
-      "short_answer_loss": NaN,
-      "step": 719,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0669,
-      "grad_norm": 1.53125,
-      "learning_rate": 9.677243663642402e-06,
-      "long_answer_loss": 0.0669,
-      "loss": 0.0948,
-      "short_answer_loss": NaN,
-      "step": 720,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0985,
-      "grad_norm": 1.421875,
-      "learning_rate": 9.645186828878875e-06,
-      "long_answer_loss": 0.0985,
-      "loss": 0.091,
-      "short_answer_loss": NaN,
-      "step": 721,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0915,
-      "grad_norm": 1.625,
-      "learning_rate": 9.613149790974748e-06,
-      "long_answer_loss": 0.0915,
-      "loss": 0.0947,
-      "short_answer_loss": NaN,
-      "step": 722,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.1125,
-      "grad_norm": 1.5390625,
-      "learning_rate": 9.581132772092642e-06,
-      "long_answer_loss": 0.1125,
-      "loss": 0.0837,
-      "short_answer_loss": NaN,
-      "step": 723,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0838,
-      "grad_norm": 1.46875,
-      "learning_rate": 9.549135994256332e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 724,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0866,
-      "grad_norm": 1.515625,
-      "learning_rate": 9.517159679349244e-06,
-      "long_answer_loss": 0.0866,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 725,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0817,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.485204049112894e-06,
-      "long_answer_loss": 0.0817,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 726,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.1044,
-      "grad_norm": 1.609375,
-      "learning_rate": 9.453269325145362e-06,
-      "long_answer_loss": 0.1044,
-      "loss": 0.0966,
-      "short_answer_loss": NaN,
-      "step": 727,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.076,
-      "grad_norm": 1.6875,
-      "learning_rate": 9.421355728899752e-06,
-      "long_answer_loss": 0.076,
-      "loss": 0.0877,
-      "short_answer_loss": NaN,
-      "step": 728,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0922,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.38946348168266e-06,
-      "long_answer_loss": 0.0922,
-      "loss": 0.0925,
-      "short_answer_loss": NaN,
-      "step": 729,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.091,
-      "grad_norm": 1.5,
-      "learning_rate": 9.357592804652636e-06,
-      "long_answer_loss": 0.091,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 730,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0953,
-      "grad_norm": 1.453125,
-      "learning_rate": 9.325743918818644e-06,
-      "long_answer_loss": 0.0953,
-      "loss": 0.089,
-      "short_answer_loss": NaN,
-      "step": 731,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0762,
-      "grad_norm": 1.578125,
-      "learning_rate": 9.293917045038538e-06,
-      "long_answer_loss": 0.0762,
-      "loss": 0.0865,
-      "short_answer_loss": NaN,
-      "step": 732,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.1159,
-      "grad_norm": 1.6484375,
-      "learning_rate": 9.262112404017536e-06,
-      "long_answer_loss": 0.1159,
-      "loss": 0.0933,
-      "short_answer_loss": NaN,
-      "step": 733,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0838,
-      "grad_norm": 1.5078125,
-      "learning_rate": 9.230330216306674e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0895,
-      "short_answer_loss": NaN,
-      "step": 734,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0787,
-      "grad_norm": 1.5078125,
-      "learning_rate": 9.198570702301276e-06,
-      "long_answer_loss": 0.0787,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 735,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0879,
-      "grad_norm": 1.5546875,
-      "learning_rate": 9.166834082239453e-06,
-      "long_answer_loss": 0.0879,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 736,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.085,
-      "grad_norm": 1.515625,
-      "learning_rate": 9.13512057620055e-06,
-      "long_answer_loss": 0.085,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 737,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0678,
-      "grad_norm": 1.5,
-      "learning_rate": 9.103430404103619e-06,
-      "long_answer_loss": 0.0678,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 738,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0784,
-      "grad_norm": 1.484375,
-      "learning_rate": 9.071763785705906e-06,
-      "long_answer_loss": 0.0784,
-      "loss": 0.0867,
-      "short_answer_loss": NaN,
-      "step": 739,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.1004,
-      "grad_norm": 1.4453125,
-      "learning_rate": 9.040120940601326e-06,
-      "long_answer_loss": 0.1004,
-      "loss": 0.0932,
-      "short_answer_loss": NaN,
-      "step": 740,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0808,
-      "grad_norm": 1.4921875,
-      "learning_rate": 9.008502088218931e-06,
-      "long_answer_loss": 0.0808,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 741,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0878,
-      "grad_norm": 1.4375,
-      "learning_rate": 8.976907447821394e-06,
-      "long_answer_loss": 0.0878,
-      "loss": 0.0871,
-      "short_answer_loss": NaN,
-      "step": 742,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.1063,
-      "grad_norm": 1.828125,
-      "learning_rate": 8.9453372385035e-06,
-      "long_answer_loss": 0.1063,
-      "loss": 0.0984,
-      "short_answer_loss": NaN,
-      "step": 743,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0968,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.913791679190604e-06,
-      "long_answer_loss": 0.0968,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 744,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0897,
-      "grad_norm": 1.5078125,
-      "learning_rate": 8.882270988637123e-06,
-      "long_answer_loss": 0.0897,
-      "loss": 0.0856,
-      "short_answer_loss": NaN,
-      "step": 745,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0761,
-      "grad_norm": 1.5078125,
-      "learning_rate": 8.85077538542503e-06,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0897,
-      "short_answer_loss": NaN,
-      "step": 746,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.093,
-      "grad_norm": 1.6171875,
-      "learning_rate": 8.819305087962316e-06,
-      "long_answer_loss": 0.093,
-      "loss": 0.0915,
-      "short_answer_loss": NaN,
-      "step": 747,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0806,
-      "grad_norm": 1.65625,
-      "learning_rate": 8.787860314481502e-06,
-      "long_answer_loss": 0.0806,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 748,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0772,
-      "grad_norm": 1.3984375,
-      "learning_rate": 8.756441283038094e-06,
-      "long_answer_loss": 0.0772,
-      "loss": 0.0854,
-      "short_answer_loss": NaN,
-      "step": 749,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0762,
-      "grad_norm": 1.453125,
-      "learning_rate": 8.725048211509113e-06,
-      "long_answer_loss": 0.0762,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 750,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0878,
-      "grad_norm": 1.5234375,
-      "learning_rate": 8.693681317591535e-06,
-      "long_answer_loss": 0.0878,
-      "loss": 0.0871,
-      "short_answer_loss": NaN,
-      "step": 751,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0957,
-      "grad_norm": 1.5625,
-      "learning_rate": 8.662340818800819e-06,
-      "long_answer_loss": 0.0957,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 752,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0688,
-      "grad_norm": 1.3828125,
-      "learning_rate": 8.631026932469383e-06,
-      "long_answer_loss": 0.0688,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 753,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0761,
-      "grad_norm": 1.5234375,
-      "learning_rate": 8.599739875745096e-06,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0885,
-      "short_answer_loss": NaN,
-      "step": 754,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0962,
-      "grad_norm": 1.4375,
-      "learning_rate": 8.56847986558978e-06,
-      "long_answer_loss": 0.0962,
-      "loss": 0.0855,
-      "short_answer_loss": NaN,
-      "step": 755,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.1052,
-      "grad_norm": 1.453125,
-      "learning_rate": 8.537247118777703e-06,
-      "long_answer_loss": 0.1052,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 756,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0654,
-      "grad_norm": 1.5625,
-      "learning_rate": 8.50604185189407e-06,
-      "long_answer_loss": 0.0654,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 757,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0946,
-      "grad_norm": 1.578125,
-      "learning_rate": 8.474864281333519e-06,
-      "long_answer_loss": 0.0946,
-      "loss": 0.0911,
-      "short_answer_loss": NaN,
-      "step": 758,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0872,
-      "grad_norm": 1.3984375,
-      "learning_rate": 8.443714623298644e-06,
-      "long_answer_loss": 0.0872,
-      "loss": 0.079,
-      "short_answer_loss": NaN,
-      "step": 759,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0908,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.412593093798457e-06,
-      "long_answer_loss": 0.0908,
-      "loss": 0.0877,
-      "short_answer_loss": NaN,
-      "step": 760,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0874,
-      "grad_norm": 1.3984375,
-      "learning_rate": 8.381499908646925e-06,
-      "long_answer_loss": 0.0874,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 761,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0735,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.350435283461452e-06,
-      "long_answer_loss": 0.0735,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 762,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0816,
-      "grad_norm": 1.5234375,
-      "learning_rate": 8.319399433661402e-06,
-      "long_answer_loss": 0.0816,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 763,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0738,
-      "grad_norm": 1.484375,
-      "learning_rate": 8.288392574466583e-06,
-      "long_answer_loss": 0.0738,
-      "loss": 0.09,
-      "short_answer_loss": NaN,
-      "step": 764,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0955,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.25741492089577e-06,
-      "long_answer_loss": 0.0955,
-      "loss": 0.0844,
-      "short_answer_loss": NaN,
-      "step": 765,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0802,
-      "grad_norm": 1.421875,
-      "learning_rate": 8.22646668776521e-06,
-      "long_answer_loss": 0.0802,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 766,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0749,
-      "grad_norm": 1.4296875,
-      "learning_rate": 8.195548089687138e-06,
-      "long_answer_loss": 0.0749,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 767,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0843,
-      "grad_norm": 1.5625,
-      "learning_rate": 8.16465934106827e-06,
-      "long_answer_loss": 0.0843,
-      "loss": 0.0995,
-      "short_answer_loss": NaN,
-      "step": 768,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0919,
-      "grad_norm": 1.46875,
-      "learning_rate": 8.13380065610834e-06,
-      "long_answer_loss": 0.0919,
-      "loss": 0.0909,
-      "short_answer_loss": NaN,
-      "step": 769,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0775,
-      "grad_norm": 1.546875,
-      "learning_rate": 8.102972248798602e-06,
-      "long_answer_loss": 0.0775,
-      "loss": 0.0862,
-      "short_answer_loss": NaN,
-      "step": 770,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0987,
-      "grad_norm": 1.4453125,
-      "learning_rate": 8.07217433292035e-06,
-      "long_answer_loss": 0.0987,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 771,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0897,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.041407122043429e-06,
-      "long_answer_loss": 0.0897,
-      "loss": 0.0954,
-      "short_answer_loss": NaN,
-      "step": 772,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.09,
-      "grad_norm": 1.46875,
-      "learning_rate": 8.010670829524753e-06,
-      "long_answer_loss": 0.09,
-      "loss": 0.0899,
-      "short_answer_loss": NaN,
-      "step": 773,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0913,
-      "grad_norm": 1.6015625,
-      "learning_rate": 7.979965668506848e-06,
-      "long_answer_loss": 0.0913,
-      "loss": 0.091,
-      "short_answer_loss": NaN,
-      "step": 774,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.1082,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.949291851916342e-06,
-      "long_answer_loss": 0.1082,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 775,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.1277,
-      "grad_norm": 1.546875,
-      "learning_rate": 7.918649592462502e-06,
-      "long_answer_loss": 0.1277,
-      "loss": 0.091,
-      "short_answer_loss": NaN,
-      "step": 776,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0676,
-      "grad_norm": 1.46875,
-      "learning_rate": 7.888039102635778e-06,
-      "long_answer_loss": 0.0676,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 777,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.1,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.857460594706296e-06,
-      "long_answer_loss": 0.1,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 778,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.1002,
-      "grad_norm": 1.5234375,
-      "learning_rate": 7.826914280722402e-06,
-      "long_answer_loss": 0.1002,
-      "loss": 0.0871,
-      "short_answer_loss": NaN,
-      "step": 779,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0749,
-      "grad_norm": 1.4609375,
-      "learning_rate": 7.796400372509199e-06,
-      "long_answer_loss": 0.0749,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 780,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.071,
-      "grad_norm": 1.3984375,
-      "learning_rate": 7.765919081667068e-06,
-      "long_answer_loss": 0.071,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 781,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0827,
-      "grad_norm": 1.4765625,
-      "learning_rate": 7.735470619570195e-06,
-      "long_answer_loss": 0.0827,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 782,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0791,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.705055197365126e-06,
-      "long_answer_loss": 0.0791,
-      "loss": 0.0852,
-      "short_answer_loss": NaN,
-      "step": 783,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0813,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.674673025969287e-06,
-      "long_answer_loss": 0.0813,
-      "loss": 0.0823,
-      "short_answer_loss": NaN,
-      "step": 784,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0715,
-      "grad_norm": 1.5,
-      "learning_rate": 7.644324316069512e-06,
-      "long_answer_loss": 0.0715,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 785,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0808,
-      "grad_norm": 1.46875,
-      "learning_rate": 7.614009278120611e-06,
-      "long_answer_loss": 0.0808,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 786,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0748,
-      "grad_norm": 1.5703125,
-      "learning_rate": 7.583728122343886e-06,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0921,
-      "short_answer_loss": NaN,
-      "step": 787,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0905,
-      "grad_norm": 1.546875,
-      "learning_rate": 7.553481058725677e-06,
-      "long_answer_loss": 0.0905,
-      "loss": 0.0895,
-      "short_answer_loss": NaN,
-      "step": 788,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0696,
-      "grad_norm": 1.515625,
-      "learning_rate": 7.523268297015916e-06,
-      "long_answer_loss": 0.0696,
-      "loss": 0.0877,
-      "short_answer_loss": NaN,
-      "step": 789,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0886,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.4930900467266755e-06,
-      "long_answer_loss": 0.0886,
-      "loss": 0.0864,
-      "short_answer_loss": NaN,
-      "step": 790,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0733,
-      "grad_norm": 1.421875,
-      "learning_rate": 7.462946517130686e-06,
-      "long_answer_loss": 0.0733,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 791,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0834,
-      "grad_norm": 1.4453125,
-      "learning_rate": 7.432837917259922e-06,
-      "long_answer_loss": 0.0834,
-      "loss": 0.0801,
-      "short_answer_loss": NaN,
-      "step": 792,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0858,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.402764455904126e-06,
-      "long_answer_loss": 0.0858,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 793,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0693,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.37272634160938e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 794,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0837,
-      "grad_norm": 1.5703125,
-      "learning_rate": 7.342723782676637e-06,
-      "long_answer_loss": 0.0837,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 795,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0897,
-      "grad_norm": 1.5078125,
-      "learning_rate": 7.312756987160299e-06,
-      "long_answer_loss": 0.0897,
-      "loss": 0.0894,
-      "short_answer_loss": NaN,
-      "step": 796,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0807,
-      "grad_norm": 1.4765625,
-      "learning_rate": 7.282826162866771e-06,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0812,
-      "short_answer_loss": NaN,
-      "step": 797,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0864,
-      "grad_norm": 1.5234375,
-      "learning_rate": 7.252931517352994e-06,
-      "long_answer_loss": 0.0864,
-      "loss": 0.092,
-      "short_answer_loss": NaN,
-      "step": 798,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0942,
-      "grad_norm": 1.40625,
-      "learning_rate": 7.223073257925047e-06,
-      "long_answer_loss": 0.0942,
-      "loss": 0.0787,
-      "short_answer_loss": NaN,
-      "step": 799,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0842,
-      "grad_norm": 1.53125,
-      "learning_rate": 7.193251591636665e-06,
-      "long_answer_loss": 0.0842,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 800,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0653,
-      "grad_norm": 1.515625,
-      "learning_rate": 7.163466725287844e-06,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 801,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0801,
-      "grad_norm": 1.4609375,
-      "learning_rate": 7.133718865423388e-06,
-      "long_answer_loss": 0.0801,
-      "loss": 0.0912,
-      "short_answer_loss": NaN,
-      "step": 802,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.1055,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.104008218331457e-06,
-      "long_answer_loss": 0.1055,
-      "loss": 0.0875,
-      "short_answer_loss": NaN,
-      "step": 803,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0916,
-      "grad_norm": 1.4296875,
-      "learning_rate": 7.074334990042189e-06,
-      "long_answer_loss": 0.0916,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 804,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0818,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.044699386326212e-06,
-      "long_answer_loss": 0.0818,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 805,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0928,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.015101612693259e-06,
-      "long_answer_loss": 0.0928,
-      "loss": 0.0891,
-      "short_answer_loss": NaN,
-      "step": 806,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0782,
-      "grad_norm": 1.484375,
-      "learning_rate": 6.985541874390725e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0847,
-      "short_answer_loss": NaN,
-      "step": 807,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0712,
-      "grad_norm": 1.5234375,
-      "learning_rate": 6.956020376402239e-06,
-      "long_answer_loss": 0.0712,
-      "loss": 0.0875,
-      "short_answer_loss": NaN,
-      "step": 808,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0711,
-      "grad_norm": 1.4375,
-      "learning_rate": 6.926537323446264e-06,
-      "long_answer_loss": 0.0711,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 809,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.081,
-      "grad_norm": 1.4921875,
-      "learning_rate": 6.897092919974652e-06,
-      "long_answer_loss": 0.081,
-      "loss": 0.0894,
-      "short_answer_loss": NaN,
-      "step": 810,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0808,
-      "grad_norm": 1.5234375,
-      "learning_rate": 6.867687370171247e-06,
-      "long_answer_loss": 0.0808,
-      "loss": 0.0906,
-      "short_answer_loss": NaN,
-      "step": 811,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0892,
-      "grad_norm": 1.46875,
-      "learning_rate": 6.838320877950458e-06,
-      "long_answer_loss": 0.0892,
-      "loss": 0.0892,
-      "short_answer_loss": NaN,
-      "step": 812,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0722,
-      "grad_norm": 1.484375,
-      "learning_rate": 6.808993646955838e-06,
-      "long_answer_loss": 0.0722,
-      "loss": 0.087,
-      "short_answer_loss": NaN,
-      "step": 813,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0721,
-      "grad_norm": 1.40625,
-      "learning_rate": 6.7797058805586926e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 814,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0864,
-      "grad_norm": 1.53125,
-      "learning_rate": 6.750457781856658e-06,
-      "long_answer_loss": 0.0864,
-      "loss": 0.0867,
-      "short_answer_loss": NaN,
-      "step": 815,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.083,
-      "grad_norm": 1.4765625,
-      "learning_rate": 6.721249553672271e-06,
-      "long_answer_loss": 0.083,
-      "loss": 0.0883,
-      "short_answer_loss": NaN,
-      "step": 816,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0823,
-      "grad_norm": 1.4609375,
-      "learning_rate": 6.6920813985516195e-06,
-      "long_answer_loss": 0.0823,
-      "loss": 0.0804,
-      "short_answer_loss": NaN,
-      "step": 817,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0912,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.6629535187628825e-06,
-      "long_answer_loss": 0.0912,
-      "loss": 0.088,
-      "short_answer_loss": NaN,
-      "step": 818,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0763,
-      "grad_norm": 1.53125,
-      "learning_rate": 6.633866116294939e-06,
-      "long_answer_loss": 0.0763,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 819,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0721,
-      "grad_norm": 1.53125,
-      "learning_rate": 6.60481939285599e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 820,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0835,
-      "grad_norm": 1.5546875,
-      "learning_rate": 6.575813549872148e-06,
-      "long_answer_loss": 0.0835,
-      "loss": 0.0957,
-      "short_answer_loss": NaN,
-      "step": 821,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0824,
-      "grad_norm": 1.5625,
-      "learning_rate": 6.546848788486021e-06,
-      "long_answer_loss": 0.0824,
-      "loss": 0.0887,
-      "short_answer_loss": NaN,
-      "step": 822,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0673,
-      "grad_norm": 1.5625,
-      "learning_rate": 6.517925309555348e-06,
-      "long_answer_loss": 0.0673,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 823,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0776,
-      "grad_norm": 1.5234375,
-      "learning_rate": 6.489043313651591e-06,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0906,
-      "short_answer_loss": NaN,
-      "step": 824,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0924,
-      "grad_norm": 1.4296875,
-      "learning_rate": 6.460203001058543e-06,
-      "long_answer_loss": 0.0924,
-      "loss": 0.0833,
-      "short_answer_loss": NaN,
-      "step": 825,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0745,
-      "grad_norm": 1.453125,
-      "learning_rate": 6.431404571770948e-06,
-      "long_answer_loss": 0.0745,
-      "loss": 0.088,
-      "short_answer_loss": NaN,
-      "step": 826,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0834,
-      "grad_norm": 1.4453125,
-      "learning_rate": 6.402648225493091e-06,
-      "long_answer_loss": 0.0834,
-      "loss": 0.09,
-      "short_answer_loss": NaN,
-      "step": 827,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0724,
-      "grad_norm": 1.4296875,
-      "learning_rate": 6.373934161637449e-06,
-      "long_answer_loss": 0.0724,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 828,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0672,
-      "grad_norm": 1.4453125,
-      "learning_rate": 6.345262579323278e-06,
-      "long_answer_loss": 0.0672,
-      "loss": 0.09,
-      "short_answer_loss": NaN,
-      "step": 829,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.1102,
-      "grad_norm": 1.4453125,
-      "learning_rate": 6.31663367737525e-06,
-      "long_answer_loss": 0.1102,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 830,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.1072,
-      "grad_norm": 1.3671875,
-      "learning_rate": 6.288047654322068e-06,
-      "long_answer_loss": 0.1072,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 831,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.1034,
-      "grad_norm": 1.40625,
-      "learning_rate": 6.259504708395078e-06,
-      "long_answer_loss": 0.1034,
-      "loss": 0.089,
-      "short_answer_loss": NaN,
-      "step": 832,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0705,
-      "grad_norm": 1.3828125,
-      "learning_rate": 6.231005037526916e-06,
-      "long_answer_loss": 0.0705,
-      "loss": 0.0857,
-      "short_answer_loss": NaN,
-      "step": 833,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.09,
-      "grad_norm": 1.4609375,
-      "learning_rate": 6.202548839350126e-06,
-      "long_answer_loss": 0.09,
-      "loss": 0.0854,
-      "short_answer_loss": NaN,
-      "step": 834,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0819,
-      "grad_norm": 1.359375,
-      "learning_rate": 6.174136311195777e-06,
-      "long_answer_loss": 0.0819,
-      "loss": 0.078,
-      "short_answer_loss": NaN,
-      "step": 835,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0914,
-      "grad_norm": 1.3984375,
-      "learning_rate": 6.1457676500921104e-06,
-      "long_answer_loss": 0.0914,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 836,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0717,
-      "grad_norm": 1.53125,
-      "learning_rate": 6.117443052763188e-06,
-      "long_answer_loss": 0.0717,
-      "loss": 0.0852,
-      "short_answer_loss": NaN,
-      "step": 837,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0772,
-      "grad_norm": 1.3828125,
-      "learning_rate": 6.089162715627474e-06,
-      "long_answer_loss": 0.0772,
-      "loss": 0.08,
-      "short_answer_loss": NaN,
-      "step": 838,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.089,
-      "grad_norm": 1.4375,
-      "learning_rate": 6.060926834796535e-06,
-      "long_answer_loss": 0.089,
-      "loss": 0.0844,
-      "short_answer_loss": NaN,
-      "step": 839,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0994,
-      "grad_norm": 1.5234375,
-      "learning_rate": 6.0327356060736475e-06,
-      "long_answer_loss": 0.0994,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 840,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0764,
-      "grad_norm": 1.3828125,
-      "learning_rate": 6.004589224952431e-06,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 841,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0858,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.976487886615526e-06,
-      "long_answer_loss": 0.0858,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 842,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0862,
-      "grad_norm": 1.484375,
-      "learning_rate": 5.94843178593321e-06,
-      "long_answer_loss": 0.0862,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 843,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0812,
-      "grad_norm": 1.5,
-      "learning_rate": 5.9204211174620645e-06,
-      "long_answer_loss": 0.0812,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 844,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0947,
-      "grad_norm": 1.46875,
-      "learning_rate": 5.892456075443616e-06,
-      "long_answer_loss": 0.0947,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 845,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0915,
-      "grad_norm": 1.4765625,
-      "learning_rate": 5.864536853802983e-06,
-      "long_answer_loss": 0.0915,
-      "loss": 0.0858,
-      "short_answer_loss": NaN,
-      "step": 846,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0862,
-      "grad_norm": 1.40625,
-      "learning_rate": 5.836663646147554e-06,
-      "long_answer_loss": 0.0862,
-      "loss": 0.0796,
-      "short_answer_loss": NaN,
-      "step": 847,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0778,
-      "grad_norm": 1.375,
-      "learning_rate": 5.808836645765628e-06,
-      "long_answer_loss": 0.0778,
-      "loss": 0.0775,
-      "short_answer_loss": NaN,
-      "step": 848,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0924,
-      "grad_norm": 1.5625,
-      "learning_rate": 5.781056045625065e-06,
-      "long_answer_loss": 0.0924,
-      "loss": 0.0926,
-      "short_answer_loss": NaN,
-      "step": 849,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0687,
-      "grad_norm": 1.4765625,
-      "learning_rate": 5.753322038371975e-06,
-      "long_answer_loss": 0.0687,
-      "loss": 0.0858,
-      "short_answer_loss": NaN,
-      "step": 850,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0741,
-      "grad_norm": 1.375,
-      "learning_rate": 5.7256348163293585e-06,
-      "long_answer_loss": 0.0741,
-      "loss": 0.0768,
-      "short_answer_loss": NaN,
-      "step": 851,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.078,
-      "grad_norm": 1.4921875,
-      "learning_rate": 5.697994571495786e-06,
-      "long_answer_loss": 0.078,
-      "loss": 0.0831,
-      "short_answer_loss": NaN,
-      "step": 852,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.093,
-      "grad_norm": 1.484375,
-      "learning_rate": 5.670401495544065e-06,
-      "long_answer_loss": 0.093,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 853,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0877,
-      "grad_norm": 1.421875,
-      "learning_rate": 5.642855779819893e-06,
-      "long_answer_loss": 0.0877,
-      "loss": 0.0885,
-      "short_answer_loss": NaN,
-      "step": 854,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0833,
-      "grad_norm": 1.453125,
-      "learning_rate": 5.615357615340558e-06,
-      "long_answer_loss": 0.0833,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 855,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.1065,
-      "grad_norm": 1.484375,
-      "learning_rate": 5.587907192793601e-06,
-      "long_answer_loss": 0.1065,
-      "loss": 0.0852,
-      "short_answer_loss": NaN,
-      "step": 856,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0855,
-      "grad_norm": 1.4296875,
-      "learning_rate": 5.560504702535489e-06,
-      "long_answer_loss": 0.0855,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 857,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0911,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.533150334590305e-06,
-      "long_answer_loss": 0.0911,
-      "loss": 0.0865,
-      "short_answer_loss": NaN,
-      "step": 858,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0909,
-      "grad_norm": 1.359375,
-      "learning_rate": 5.505844278648424e-06,
-      "long_answer_loss": 0.0909,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 859,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0901,
-      "grad_norm": 1.5078125,
-      "learning_rate": 5.47858672406519e-06,
-      "long_answer_loss": 0.0901,
-      "loss": 0.0878,
-      "short_answer_loss": NaN,
-      "step": 860,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0916,
-      "grad_norm": 1.6015625,
-      "learning_rate": 5.451377859859623e-06,
-      "long_answer_loss": 0.0916,
-      "loss": 0.0936,
-      "short_answer_loss": NaN,
-      "step": 861,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0748,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.424217874713096e-06,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0841,
-      "short_answer_loss": NaN,
-      "step": 862,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0997,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.397106956968012e-06,
-      "long_answer_loss": 0.0997,
-      "loss": 0.0795,
-      "short_answer_loss": NaN,
-      "step": 863,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.072,
-      "grad_norm": 1.5234375,
-      "learning_rate": 5.370045294626544e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0854,
-      "short_answer_loss": NaN,
-      "step": 864,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.095,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.343033075349266e-06,
-      "long_answer_loss": 0.095,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 865,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.09,
-      "grad_norm": 1.40625,
-      "learning_rate": 5.31607048645391e-06,
-      "long_answer_loss": 0.09,
-      "loss": 0.0806,
-      "short_answer_loss": NaN,
-      "step": 866,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0883,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.289157714914039e-06,
-      "long_answer_loss": 0.0883,
-      "loss": 0.0823,
-      "short_answer_loss": NaN,
-      "step": 867,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0823,
-      "grad_norm": 1.3984375,
-      "learning_rate": 5.262294947357744e-06,
-      "long_answer_loss": 0.0823,
-      "loss": 0.0808,
-      "short_answer_loss": NaN,
-      "step": 868,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0869,
-      "grad_norm": 1.546875,
-      "learning_rate": 5.235482370066372e-06,
-      "long_answer_loss": 0.0869,
-      "loss": 0.0864,
-      "short_answer_loss": NaN,
-      "step": 869,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0882,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.208720168973219e-06,
-      "long_answer_loss": 0.0882,
-      "loss": 0.0884,
-      "short_answer_loss": NaN,
-      "step": 870,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0815,
-      "grad_norm": 1.3984375,
-      "learning_rate": 5.1820085296622455e-06,
-      "long_answer_loss": 0.0815,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 871,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0988,
-      "grad_norm": 1.4296875,
-      "learning_rate": 5.15534763736679e-06,
-      "long_answer_loss": 0.0988,
-      "loss": 0.08,
-      "short_answer_loss": NaN,
-      "step": 872,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0807,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.128737676968274e-06,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0867,
-      "short_answer_loss": NaN,
-      "step": 873,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0773,
-      "grad_norm": 1.421875,
-      "learning_rate": 5.102178832994941e-06,
-      "long_answer_loss": 0.0773,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 874,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0802,
-      "grad_norm": 1.4609375,
-      "learning_rate": 5.0756712896205625e-06,
-      "long_answer_loss": 0.0802,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 875,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0838,
-      "grad_norm": 1.4609375,
-      "learning_rate": 5.049215230663152e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 876,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0724,
-      "grad_norm": 1.390625,
-      "learning_rate": 5.022810839583707e-06,
-      "long_answer_loss": 0.0724,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 877,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.1062,
-      "grad_norm": 1.5390625,
-      "learning_rate": 4.99645829948495e-06,
-      "long_answer_loss": 0.1062,
-      "loss": 0.0844,
-      "short_answer_loss": NaN,
-      "step": 878,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0807,
-      "grad_norm": 1.546875,
-      "learning_rate": 4.970157793110011e-06,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0853,
-      "short_answer_loss": NaN,
-      "step": 879,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.077,
-      "grad_norm": 1.40625,
-      "learning_rate": 4.9439095028412055e-06,
-      "long_answer_loss": 0.077,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 880,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0782,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.917713610698754e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0853,
-      "short_answer_loss": NaN,
-      "step": 881,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0888,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.891570298339508e-06,
-      "long_answer_loss": 0.0888,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 882,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0763,
-      "grad_norm": 1.390625,
-      "learning_rate": 4.8654797470557125e-06,
-      "long_answer_loss": 0.0763,
-      "loss": 0.0803,
-      "short_answer_loss": NaN,
-      "step": 883,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0933,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.8394421377737354e-06,
-      "long_answer_loss": 0.0933,
-      "loss": 0.0853,
-      "short_answer_loss": NaN,
-      "step": 884,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0856,
-      "grad_norm": 1.5234375,
-      "learning_rate": 4.813457651052815e-06,
-      "long_answer_loss": 0.0856,
-      "loss": 0.0894,
-      "short_answer_loss": NaN,
-      "step": 885,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0769,
-      "grad_norm": 1.34375,
-      "learning_rate": 4.7875264670838096e-06,
-      "long_answer_loss": 0.0769,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 886,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.072,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.76164876568794e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 887,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0695,
-      "grad_norm": 1.5234375,
-      "learning_rate": 4.735824726315555e-06,
-      "long_answer_loss": 0.0695,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 888,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0804,
-      "grad_norm": 1.4921875,
-      "learning_rate": 4.710054528044885e-06,
-      "long_answer_loss": 0.0804,
-      "loss": 0.0894,
-      "short_answer_loss": NaN,
-      "step": 889,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0988,
-      "grad_norm": 1.4921875,
-      "learning_rate": 4.68433834958078e-06,
-      "long_answer_loss": 0.0988,
-      "loss": 0.0898,
-      "short_answer_loss": NaN,
-      "step": 890,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.084,
-      "grad_norm": 1.4765625,
-      "learning_rate": 4.658676369253513e-06,
-      "long_answer_loss": 0.084,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 891,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0776,
-      "grad_norm": 1.3359375,
-      "learning_rate": 4.633068765017493e-06,
-      "long_answer_loss": 0.0776,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 892,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0767,
-      "grad_norm": 1.5703125,
-      "learning_rate": 4.607515714450064e-06,
-      "long_answer_loss": 0.0767,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 893,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0933,
-      "grad_norm": 1.3515625,
-      "learning_rate": 4.582017394750271e-06,
-      "long_answer_loss": 0.0933,
-      "loss": 0.0855,
-      "short_answer_loss": NaN,
-      "step": 894,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0737,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.556573982737611e-06,
-      "long_answer_loss": 0.0737,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 895,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0906,
-      "grad_norm": 1.3984375,
-      "learning_rate": 4.531185654850829e-06,
-      "long_answer_loss": 0.0906,
-      "loss": 0.0856,
-      "short_answer_loss": NaN,
-      "step": 896,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.1176,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.505852587146689e-06,
-      "long_answer_loss": 0.1176,
-      "loss": 0.0853,
-      "short_answer_loss": NaN,
-      "step": 897,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0898,
-      "grad_norm": 1.4453125,
-      "learning_rate": 4.480574955298746e-06,
-      "long_answer_loss": 0.0898,
-      "loss": 0.0841,
-      "short_answer_loss": NaN,
-      "step": 898,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0729,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.4553529345961285e-06,
-      "long_answer_loss": 0.0729,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 899,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0803,
-      "grad_norm": 1.34375,
-      "learning_rate": 4.430186699942337e-06,
-      "long_answer_loss": 0.0803,
-      "loss": 0.0797,
-      "short_answer_loss": NaN,
-      "step": 900,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.072,
-      "grad_norm": 1.3515625,
-      "learning_rate": 4.405076425854007e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0785,
-      "short_answer_loss": NaN,
-      "step": 901,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0764,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.380022286459727e-06,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0803,
-      "short_answer_loss": NaN,
-      "step": 902,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0782,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.3550244554987975e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 903,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0868,
-      "grad_norm": 1.484375,
-      "learning_rate": 4.33008310632006e-06,
-      "long_answer_loss": 0.0868,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 904,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0722,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.305198411880686e-06,
-      "long_answer_loss": 0.0722,
-      "loss": 0.084,
-      "short_answer_loss": NaN,
-      "step": 905,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0877,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.280370544744949e-06,
-      "long_answer_loss": 0.0877,
-      "loss": 0.0865,
-      "short_answer_loss": NaN,
-      "step": 906,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0978,
-      "grad_norm": 1.46875,
-      "learning_rate": 4.25559967708307e-06,
-      "long_answer_loss": 0.0978,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 907,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0803,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.230885980669999e-06,
-      "long_answer_loss": 0.0803,
-      "loss": 0.0876,
-      "short_answer_loss": NaN,
-      "step": 908,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.1023,
-      "grad_norm": 1.5,
-      "learning_rate": 4.2062296268842196e-06,
-      "long_answer_loss": 0.1023,
-      "loss": 0.0847,
-      "short_answer_loss": NaN,
-      "step": 909,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0721,
-      "grad_norm": 1.3984375,
-      "learning_rate": 4.181630786706582e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 910,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.106,
-      "grad_norm": 1.5,
-      "learning_rate": 4.157089630719099e-06,
-      "long_answer_loss": 0.106,
-      "loss": 0.0985,
-      "short_answer_loss": NaN,
-      "step": 911,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0823,
-      "grad_norm": 1.375,
-      "learning_rate": 4.1326063291037705e-06,
-      "long_answer_loss": 0.0823,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 912,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0964,
-      "grad_norm": 1.5078125,
-      "learning_rate": 4.108181051641408e-06,
-      "long_answer_loss": 0.0964,
-      "loss": 0.0892,
-      "short_answer_loss": NaN,
-      "step": 913,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0932,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.083813967710435e-06,
-      "long_answer_loss": 0.0932,
-      "loss": 0.0842,
-      "short_answer_loss": NaN,
-      "step": 914,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0904,
-      "grad_norm": 1.5390625,
-      "learning_rate": 4.059505246285743e-06,
-      "long_answer_loss": 0.0904,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 915,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0775,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.035255055937507e-06,
-      "long_answer_loss": 0.0775,
-      "loss": 0.0779,
-      "short_answer_loss": NaN,
-      "step": 916,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0812,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.011063564829997e-06,
-      "long_answer_loss": 0.0812,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 917,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0924,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.986930940720461e-06,
-      "long_answer_loss": 0.0924,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 918,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.071,
-      "grad_norm": 1.375,
-      "learning_rate": 3.962857350957896e-06,
-      "long_answer_loss": 0.071,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 919,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0876,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.938842962481945e-06,
-      "long_answer_loss": 0.0876,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 920,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0865,
-      "grad_norm": 1.4375,
-      "learning_rate": 3.9148879418217115e-06,
-      "long_answer_loss": 0.0865,
-      "loss": 0.0901,
-      "short_answer_loss": NaN,
-      "step": 921,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.08,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.890992455094601e-06,
-      "long_answer_loss": 0.08,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 922,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0704,
-      "grad_norm": 1.46875,
-      "learning_rate": 3.867156668005188e-06,
-      "long_answer_loss": 0.0704,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 923,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0806,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.8433807458440525e-06,
-      "long_answer_loss": 0.0806,
-      "loss": 0.0767,
-      "short_answer_loss": NaN,
-      "step": 924,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0928,
-      "grad_norm": 1.5625,
-      "learning_rate": 3.8196648534866386e-06,
-      "long_answer_loss": 0.0928,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 925,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0807,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.7960091553921094e-06,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 926,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0838,
-      "grad_norm": 1.515625,
-      "learning_rate": 3.7724138156022132e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 927,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.09,
-      "grad_norm": 1.5078125,
-      "learning_rate": 3.7488789977401282e-06,
-      "long_answer_loss": 0.09,
-      "loss": 0.0887,
-      "short_answer_loss": NaN,
-      "step": 928,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.087,
-      "grad_norm": 1.5078125,
-      "learning_rate": 3.725404865009348e-06,
-      "long_answer_loss": 0.087,
-      "loss": 0.0896,
-      "short_answer_loss": NaN,
-      "step": 929,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0751,
-      "grad_norm": 1.3671875,
-      "learning_rate": 3.7019915801925474e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 930,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0687,
-      "grad_norm": 1.328125,
-      "learning_rate": 3.6786393056504277e-06,
-      "long_answer_loss": 0.0687,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 931,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.089,
-      "grad_norm": 1.4765625,
-      "learning_rate": 3.6553482033206422e-06,
-      "long_answer_loss": 0.089,
-      "loss": 0.0932,
-      "short_answer_loss": NaN,
-      "step": 932,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.1103,
-      "grad_norm": 1.5078125,
-      "learning_rate": 3.6321184347166124e-06,
-      "long_answer_loss": 0.1103,
-      "loss": 0.0907,
-      "short_answer_loss": NaN,
-      "step": 933,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.092,
-      "grad_norm": 1.6875,
-      "learning_rate": 3.608950160926454e-06,
-      "long_answer_loss": 0.092,
-      "loss": 0.0808,
-      "short_answer_loss": NaN,
-      "step": 934,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0665,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.5858435426118426e-06,
-      "long_answer_loss": 0.0665,
-      "loss": 0.0768,
-      "short_answer_loss": NaN,
-      "step": 935,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0871,
-      "grad_norm": 1.3671875,
-      "learning_rate": 3.5627987400068924e-06,
-      "long_answer_loss": 0.0871,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 936,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0774,
-      "grad_norm": 1.53125,
-      "learning_rate": 3.5398159129170577e-06,
-      "long_answer_loss": 0.0774,
-      "loss": 0.0856,
-      "short_answer_loss": NaN,
-      "step": 937,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0776,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.516895220718025e-06,
-      "long_answer_loss": 0.0776,
-      "loss": 0.08,
-      "short_answer_loss": NaN,
-      "step": 938,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0764,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.4940368223545975e-06,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 939,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0876,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.4712408763396013e-06,
-      "long_answer_loss": 0.0876,
-      "loss": 0.0881,
-      "short_answer_loss": NaN,
-      "step": 940,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0918,
-      "grad_norm": 1.4375,
-      "learning_rate": 3.4485075407527755e-06,
-      "long_answer_loss": 0.0918,
-      "loss": 0.0906,
-      "short_answer_loss": NaN,
-      "step": 941,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0601,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.4258369732396934e-06,
-      "long_answer_loss": 0.0601,
-      "loss": 0.0804,
-      "short_answer_loss": NaN,
-      "step": 942,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0877,
-      "grad_norm": 1.4765625,
-      "learning_rate": 3.40322933101066e-06,
-      "long_answer_loss": 0.0877,
-      "loss": 0.0827,
-      "short_answer_loss": NaN,
-      "step": 943,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0845,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.3806847708396054e-06,
-      "long_answer_loss": 0.0845,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 944,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0625,
-      "grad_norm": 1.3671875,
-      "learning_rate": 3.3582034490630366e-06,
-      "long_answer_loss": 0.0625,
-      "loss": 0.0862,
-      "short_answer_loss": NaN,
-      "step": 945,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0842,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.3357855215789214e-06,
-      "long_answer_loss": 0.0842,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 946,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0917,
-      "grad_norm": 1.4140625,
-      "learning_rate": 3.3134311438456064e-06,
-      "long_answer_loss": 0.0917,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 947,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0973,
-      "grad_norm": 1.3828125,
-      "learning_rate": 3.291140470880759e-06,
-      "long_answer_loss": 0.0973,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 948,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0791,
-      "grad_norm": 1.5859375,
-      "learning_rate": 3.268913657260285e-06,
-      "long_answer_loss": 0.0791,
-      "loss": 0.0956,
-      "short_answer_loss": NaN,
-      "step": 949,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0989,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.2467508571172437e-06,
-      "long_answer_loss": 0.0989,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 950,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.1028,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.2246522241407965e-06,
-      "long_answer_loss": 0.1028,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 951,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0698,
-      "grad_norm": 1.46875,
-      "learning_rate": 3.2026179115751344e-06,
-      "long_answer_loss": 0.0698,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 952,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0667,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.1806480722184134e-06,
-      "long_answer_loss": 0.0667,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 953,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0943,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.1587428584216998e-06,
-      "long_answer_loss": 0.0943,
-      "loss": 0.0903,
-      "short_answer_loss": NaN,
-      "step": 954,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0777,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.136902422087898e-06,
-      "long_answer_loss": 0.0777,
-      "loss": 0.0794,
-      "short_answer_loss": NaN,
-      "step": 955,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0694,
-      "grad_norm": 1.328125,
-      "learning_rate": 3.115126914670724e-06,
-      "long_answer_loss": 0.0694,
-      "loss": 0.0762,
-      "short_answer_loss": NaN,
-      "step": 956,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0919,
-      "grad_norm": 1.4375,
-      "learning_rate": 3.093416487173638e-06,
-      "long_answer_loss": 0.0919,
-      "loss": 0.0886,
-      "short_answer_loss": NaN,
-      "step": 957,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0752,
-      "grad_norm": 1.3828125,
-      "learning_rate": 3.071771290148788e-06,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 958,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.076,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.050191473695997e-06,
-      "long_answer_loss": 0.076,
-      "loss": 0.0844,
-      "short_answer_loss": NaN,
-      "step": 959,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.1016,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.028677187461687e-06,
-      "long_answer_loss": 0.1016,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 960,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0936,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.007228580637862e-06,
-      "long_answer_loss": 0.0936,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 961,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0809,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.985845801961075e-06,
-      "long_answer_loss": 0.0809,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 962,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0871,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.964528999711376e-06,
-      "long_answer_loss": 0.0871,
-      "loss": 0.0808,
-      "short_answer_loss": NaN,
-      "step": 963,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0838,
-      "grad_norm": 1.375,
-      "learning_rate": 2.9432783217113055e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 964,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.099,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.922093915324875e-06,
-      "long_answer_loss": 0.099,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 965,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0942,
-      "grad_norm": 1.375,
-      "learning_rate": 2.900975927456508e-06,
-      "long_answer_loss": 0.0942,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 966,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0817,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.879924504550058e-06,
-      "long_answer_loss": 0.0817,
-      "loss": 0.0899,
-      "short_answer_loss": NaN,
-      "step": 967,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0851,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.858939792587782e-06,
-      "long_answer_loss": 0.0851,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 968,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.069,
-      "grad_norm": 1.3828125,
-      "learning_rate": 2.8380219370893192e-06,
-      "long_answer_loss": 0.069,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 969,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0749,
-      "grad_norm": 1.3359375,
-      "learning_rate": 2.817171083110691e-06,
-      "long_answer_loss": 0.0749,
-      "loss": 0.081,
-      "short_answer_loss": NaN,
-      "step": 970,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.096,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.796387375243298e-06,
-      "long_answer_loss": 0.096,
-      "loss": 0.087,
-      "short_answer_loss": NaN,
-      "step": 971,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0782,
-      "grad_norm": 1.390625,
-      "learning_rate": 2.775670957612908e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 972,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0783,
-      "grad_norm": 1.5234375,
-      "learning_rate": 2.755021973878663e-06,
-      "long_answer_loss": 0.0783,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 973,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0794,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.734440567232077e-06,
-      "long_answer_loss": 0.0794,
-      "loss": 0.0787,
-      "short_answer_loss": NaN,
-      "step": 974,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.09,
-      "grad_norm": 1.4453125,
-      "learning_rate": 2.713926880396049e-06,
-      "long_answer_loss": 0.09,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 975,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0732,
-      "grad_norm": 1.4921875,
-      "learning_rate": 2.693481055623878e-06,
-      "long_answer_loss": 0.0732,
-      "loss": 0.089,
-      "short_answer_loss": NaN,
-      "step": 976,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0841,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.673103234698256e-06,
-      "long_answer_loss": 0.0841,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 977,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.1165,
-      "grad_norm": 1.5625,
-      "learning_rate": 2.65279355893031e-06,
-      "long_answer_loss": 0.1165,
-      "loss": 0.0844,
-      "short_answer_loss": NaN,
-      "step": 978,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0704,
-      "grad_norm": 1.359375,
-      "learning_rate": 2.6325521691586057e-06,
-      "long_answer_loss": 0.0704,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 979,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0962,
-      "grad_norm": 1.484375,
-      "learning_rate": 2.612379205748178e-06,
-      "long_answer_loss": 0.0962,
-      "loss": 0.0843,
-      "short_answer_loss": NaN,
-      "step": 980,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0906,
-      "grad_norm": 1.5546875,
-      "learning_rate": 2.592274808589558e-06,
-      "long_answer_loss": 0.0906,
-      "loss": 0.0907,
-      "short_answer_loss": NaN,
-      "step": 981,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0914,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.5722391170977896e-06,
-      "long_answer_loss": 0.0914,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 982,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.113,
-      "grad_norm": 1.375,
-      "learning_rate": 2.552272270211484e-06,
-      "long_answer_loss": 0.113,
-      "loss": 0.0828,
-      "short_answer_loss": NaN,
-      "step": 983,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0792,
-      "grad_norm": 1.5390625,
-      "learning_rate": 2.5323744063918423e-06,
-      "long_answer_loss": 0.0792,
-      "loss": 0.0906,
-      "short_answer_loss": NaN,
-      "step": 984,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.1093,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.5125456636216987e-06,
-      "long_answer_loss": 0.1093,
-      "loss": 0.0864,
-      "short_answer_loss": NaN,
-      "step": 985,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0726,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.4927861794045633e-06,
-      "long_answer_loss": 0.0726,
-      "loss": 0.0825,
-      "short_answer_loss": NaN,
-      "step": 986,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.078,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.473096090763674e-06,
-      "long_answer_loss": 0.078,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 987,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0892,
-      "grad_norm": 1.3828125,
-      "learning_rate": 2.4534755342410286e-06,
-      "long_answer_loss": 0.0892,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 988,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0843,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.433924645896464e-06,
-      "long_answer_loss": 0.0843,
-      "loss": 0.0791,
-      "short_answer_loss": NaN,
-      "step": 989,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0857,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.4144435613066938e-06,
-      "long_answer_loss": 0.0857,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 990,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0943,
-      "grad_norm": 1.5,
-      "learning_rate": 2.395032415564366e-06,
-      "long_answer_loss": 0.0943,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 991,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0844,
-      "grad_norm": 1.4921875,
-      "learning_rate": 2.3756913432771547e-06,
-      "long_answer_loss": 0.0844,
-      "loss": 0.0887,
-      "short_answer_loss": NaN,
-      "step": 992,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0862,
-      "grad_norm": 1.359375,
-      "learning_rate": 2.3564204785667815e-06,
-      "long_answer_loss": 0.0862,
-      "loss": 0.0791,
-      "short_answer_loss": NaN,
-      "step": 993,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0689,
-      "grad_norm": 1.484375,
-      "learning_rate": 2.3372199550681263e-06,
-      "long_answer_loss": 0.0689,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 994,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0752,
-      "grad_norm": 1.34375,
-      "learning_rate": 2.318089905928278e-06,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0774,
-      "short_answer_loss": NaN,
-      "step": 995,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0785,
-      "grad_norm": 1.484375,
-      "learning_rate": 2.299030463805614e-06,
-      "long_answer_loss": 0.0785,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 996,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0745,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.2800417608688894e-06,
-      "long_answer_loss": 0.0745,
-      "loss": 0.0804,
-      "short_answer_loss": NaN,
-      "step": 997,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0745,
-      "grad_norm": 1.4453125,
-      "learning_rate": 2.2611239287963133e-06,
-      "long_answer_loss": 0.0745,
-      "loss": 0.0836,
-      "short_answer_loss": NaN,
-      "step": 998,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0708,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.242277098774634e-06,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 999,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0711,
-      "grad_norm": 1.3828125,
-      "learning_rate": 2.2235014014982374e-06,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0794,
-      "short_answer_loss": NaN,
-      "step": 1000,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0783,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.204796967168225e-06,
-      "long_answer_loss": 0.0783,
-      "loss": 0.0788,
-      "short_answer_loss": NaN,
-      "step": 1001,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0967,
-      "grad_norm": 1.359375,
-      "learning_rate": 2.1861639254915343e-06,
-      "long_answer_loss": 0.0967,
-      "loss": 0.078,
-      "short_answer_loss": NaN,
-      "step": 1002,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0626,
-      "grad_norm": 1.40625,
-      "learning_rate": 2.167602405680021e-06,
-      "long_answer_loss": 0.0626,
-      "loss": 0.0861,
-      "short_answer_loss": NaN,
-      "step": 1003,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.1006,
-      "grad_norm": 1.5078125,
-      "learning_rate": 2.1491125364495615e-06,
-      "long_answer_loss": 0.1006,
-      "loss": 0.0877,
-      "short_answer_loss": NaN,
-      "step": 1004,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0726,
-      "grad_norm": 1.375,
-      "learning_rate": 2.130694446019177e-06,
-      "long_answer_loss": 0.0726,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 1005,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0736,
-      "grad_norm": 1.4453125,
-      "learning_rate": 2.112348262110138e-06,
-      "long_answer_loss": 0.0736,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 1006,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.1027,
-      "grad_norm": 1.484375,
-      "learning_rate": 2.0940741119450615e-06,
-      "long_answer_loss": 0.1027,
-      "loss": 0.0941,
-      "short_answer_loss": NaN,
-      "step": 1007,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.1083,
-      "grad_norm": 1.375,
-      "learning_rate": 2.075872122247051e-06,
-      "long_answer_loss": 0.1083,
-      "loss": 0.0809,
-      "short_answer_loss": NaN,
-      "step": 1008,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0752,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.0577424192388117e-06,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 1009,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0827,
-      "grad_norm": 1.40625,
-      "learning_rate": 2.0396851286417647e-06,
-      "long_answer_loss": 0.0827,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 1010,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0947,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.021700375675188e-06,
-      "long_answer_loss": 0.0947,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 1011,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0824,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.0037882850553455e-06,
-      "long_answer_loss": 0.0824,
-      "loss": 0.085,
-      "short_answer_loss": NaN,
-      "step": 1012,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.063,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.9859489809946176e-06,
-      "long_answer_loss": 0.063,
-      "loss": 0.081,
-      "short_answer_loss": NaN,
-      "step": 1013,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0673,
-      "grad_norm": 1.3203125,
-      "learning_rate": 1.9681825872006427e-06,
-      "long_answer_loss": 0.0673,
-      "loss": 0.075,
-      "short_answer_loss": NaN,
-      "step": 1014,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0714,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.950489226875454e-06,
-      "long_answer_loss": 0.0714,
-      "loss": 0.0838,
-      "short_answer_loss": NaN,
-      "step": 1015,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0823,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.9328690227146394e-06,
-      "long_answer_loss": 0.0823,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 1016,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0778,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.9153220969064788e-06,
-      "long_answer_loss": 0.0778,
-      "loss": 0.0786,
-      "short_answer_loss": NaN,
-      "step": 1017,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0987,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.8978485711310862e-06,
-      "long_answer_loss": 0.0987,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 1018,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.1099,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.8804485665596033e-06,
-      "long_answer_loss": 0.1099,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 1019,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0835,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.8631222038533124e-06,
-      "long_answer_loss": 0.0835,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 1020,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0782,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.845869603162835e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 1021,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0854,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.8286908841272834e-06,
-      "long_answer_loss": 0.0854,
-      "loss": 0.0889,
-      "short_answer_loss": NaN,
-      "step": 1022,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0844,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.8115861658734287e-06,
-      "long_answer_loss": 0.0844,
-      "loss": 0.0833,
-      "short_answer_loss": NaN,
-      "step": 1023,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0728,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.7945555670148856e-06,
-      "long_answer_loss": 0.0728,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 1024,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0882,
-      "grad_norm": 1.3046875,
-      "learning_rate": 1.77759920565128e-06,
-      "long_answer_loss": 0.0882,
-      "loss": 0.0843,
-      "short_answer_loss": NaN,
-      "step": 1025,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0867,
-      "grad_norm": 1.359375,
-      "learning_rate": 1.7607171993674371e-06,
-      "long_answer_loss": 0.0867,
-      "loss": 0.0762,
-      "short_answer_loss": NaN,
-      "step": 1026,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.079,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.7439096652325584e-06,
-      "long_answer_loss": 0.079,
-      "loss": 0.0806,
-      "short_answer_loss": NaN,
-      "step": 1027,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.092,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.7271767197994213e-06,
-      "long_answer_loss": 0.092,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 1028,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0865,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.710518479103551e-06,
-      "long_answer_loss": 0.0865,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 1029,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0924,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.693935058662445e-06,
-      "long_answer_loss": 0.0924,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 1030,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0865,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.6774265734747384e-06,
-      "long_answer_loss": 0.0865,
-      "loss": 0.0739,
-      "short_answer_loss": NaN,
-      "step": 1031,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.1034,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.6609931380194358e-06,
-      "long_answer_loss": 0.1034,
-      "loss": 0.0891,
-      "short_answer_loss": NaN,
-      "step": 1032,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0842,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.644634866255107e-06,
-      "long_answer_loss": 0.0842,
-      "loss": 0.0892,
-      "short_answer_loss": NaN,
-      "step": 1033,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0779,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.628351871619084e-06,
-      "long_answer_loss": 0.0779,
-      "loss": 0.0773,
-      "short_answer_loss": NaN,
-      "step": 1034,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0844,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.6121442670266925e-06,
-      "long_answer_loss": 0.0844,
-      "loss": 0.0852,
-      "short_answer_loss": NaN,
-      "step": 1035,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.1114,
-      "grad_norm": 1.3359375,
-      "learning_rate": 1.5960121648704626e-06,
-      "long_answer_loss": 0.1114,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 1036,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0986,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.579955677019343e-06,
-      "long_answer_loss": 0.0986,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 1037,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0789,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.5639749148179368e-06,
-      "long_answer_loss": 0.0789,
-      "loss": 0.0812,
-      "short_answer_loss": NaN,
-      "step": 1038,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.087,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.548069989085721e-06,
-      "long_answer_loss": 0.087,
-      "loss": 0.0795,
-      "short_answer_loss": NaN,
-      "step": 1039,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0908,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.5322410101162796e-06,
-      "long_answer_loss": 0.0908,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 1040,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0962,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.516488087676543e-06,
-      "long_answer_loss": 0.0962,
-      "loss": 0.0874,
-      "short_answer_loss": NaN,
-      "step": 1041,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0835,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.5008113310060148e-06,
-      "long_answer_loss": 0.0835,
-      "loss": 0.0855,
-      "short_answer_loss": NaN,
-      "step": 1042,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0724,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.4852108488160349e-06,
-      "long_answer_loss": 0.0724,
-      "loss": 0.0871,
-      "short_answer_loss": NaN,
-      "step": 1043,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0869,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.4696867492890066e-06,
-      "long_answer_loss": 0.0869,
-      "loss": 0.0887,
-      "short_answer_loss": NaN,
-      "step": 1044,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0802,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.4542391400776484e-06,
-      "long_answer_loss": 0.0802,
-      "loss": 0.0779,
-      "short_answer_loss": NaN,
-      "step": 1045,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0871,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.4388681283042675e-06,
-      "long_answer_loss": 0.0871,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 1046,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0739,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.423573820559987e-06,
-      "long_answer_loss": 0.0739,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 1047,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0827,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.408356322904028e-06,
-      "long_answer_loss": 0.0827,
-      "loss": 0.077,
-      "short_answer_loss": NaN,
-      "step": 1048,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0925,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.3932157408629715e-06,
-      "long_answer_loss": 0.0925,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 1049,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0879,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.3781521794300145e-06,
-      "long_answer_loss": 0.0879,
-      "loss": 0.0884,
-      "short_answer_loss": NaN,
-      "step": 1050,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0786,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.363165743064254e-06,
-      "long_answer_loss": 0.0786,
-      "loss": 0.0948,
-      "short_answer_loss": NaN,
-      "step": 1051,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.098,
-      "grad_norm": 1.3203125,
-      "learning_rate": 1.3482565356899637e-06,
-      "long_answer_loss": 0.098,
-      "loss": 0.0819,
-      "short_answer_loss": NaN,
-      "step": 1052,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0926,
-      "grad_norm": 1.375,
-      "learning_rate": 1.3334246606958616e-06,
-      "long_answer_loss": 0.0926,
-      "loss": 0.0767,
-      "short_answer_loss": NaN,
-      "step": 1053,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0933,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.3186702209344052e-06,
-      "long_answer_loss": 0.0933,
-      "loss": 0.0835,
-      "short_answer_loss": NaN,
-      "step": 1054,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0866,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.303993318721071e-06,
-      "long_answer_loss": 0.0866,
-      "loss": 0.0773,
-      "short_answer_loss": NaN,
-      "step": 1055,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0888,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.2893940558336427e-06,
-      "long_answer_loss": 0.0888,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 1056,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.068,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.2748725335115175e-06,
-      "long_answer_loss": 0.068,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 1057,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0896,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.260428852454995e-06,
-      "long_answer_loss": 0.0896,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 1058,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0721,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.2460631128245705e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 1059,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0784,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2317754142402674e-06,
-      "long_answer_loss": 0.0784,
-      "loss": 0.0897,
-      "short_answer_loss": NaN,
-      "step": 1060,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.077,
-      "grad_norm": 1.5,
-      "learning_rate": 1.2175658557809139e-06,
-      "long_answer_loss": 0.077,
-      "loss": 0.0825,
-      "short_answer_loss": NaN,
-      "step": 1061,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.1099,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.2034345359834775e-06,
-      "long_answer_loss": 0.1099,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 1062,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0767,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.1893815528423793e-06,
-      "long_answer_loss": 0.0767,
-      "loss": 0.0823,
-      "short_answer_loss": NaN,
-      "step": 1063,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0766,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.1754070038087984e-06,
-      "long_answer_loss": 0.0766,
-      "loss": 0.0871,
-      "short_answer_loss": NaN,
-      "step": 1064,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.1005,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.1615109857900208e-06,
-      "long_answer_loss": 0.1005,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 1065,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0747,
-      "grad_norm": 1.3359375,
-      "learning_rate": 1.1476935951487494e-06,
-      "long_answer_loss": 0.0747,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 1066,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0735,
-      "grad_norm": 1.359375,
-      "learning_rate": 1.1339549277024408e-06,
-      "long_answer_loss": 0.0735,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 1067,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.115,
-      "grad_norm": 1.34375,
-      "learning_rate": 1.1202950787226453e-06,
-      "long_answer_loss": 0.115,
-      "loss": 0.0882,
-      "short_answer_loss": NaN,
-      "step": 1068,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0886,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.1067141429343356e-06,
-      "long_answer_loss": 0.0886,
-      "loss": 0.0866,
-      "short_answer_loss": NaN,
-      "step": 1069,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0841,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.0932122145152601e-06,
-      "long_answer_loss": 0.0841,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 1070,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0916,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.0797893870952897e-06,
-      "long_answer_loss": 0.0916,
-      "loss": 0.0862,
-      "short_answer_loss": NaN,
-      "step": 1071,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0967,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.0664457537557543e-06,
-      "long_answer_loss": 0.0967,
-      "loss": 0.0924,
-      "short_answer_loss": NaN,
-      "step": 1072,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0863,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.0531814070288234e-06,
-      "long_answer_loss": 0.0863,
-      "loss": 0.0791,
-      "short_answer_loss": NaN,
-      "step": 1073,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0847,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.0399964388968397e-06,
-      "long_answer_loss": 0.0847,
-      "loss": 0.0862,
-      "short_answer_loss": NaN,
-      "step": 1074,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.112,
-      "grad_norm": 1.90625,
-      "learning_rate": 1.0268909407916874e-06,
-      "long_answer_loss": 0.112,
-      "loss": 0.0878,
-      "short_answer_loss": NaN,
-      "step": 1075,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0947,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.0138650035941694e-06,
-      "long_answer_loss": 0.0947,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 1076,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0711,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.0009187176333672e-06,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 1077,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.1091,
-      "grad_norm": 1.4453125,
-      "learning_rate": 9.880521726860103e-07,
-      "long_answer_loss": 0.1091,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 1078,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.1005,
-      "grad_norm": 1.4296875,
-      "learning_rate": 9.752654579758672e-07,
-      "long_answer_loss": 0.1005,
-      "loss": 0.087,
-      "short_answer_loss": NaN,
-      "step": 1079,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0745,
-      "grad_norm": 1.34375,
-      "learning_rate": 9.625586621731158e-07,
-      "long_answer_loss": 0.0745,
-      "loss": 0.0795,
-      "short_answer_loss": NaN,
-      "step": 1080,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0781,
-      "grad_norm": 1.4296875,
-      "learning_rate": 9.499318733937334e-07,
-      "long_answer_loss": 0.0781,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 1081,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0762,
-      "grad_norm": 1.390625,
-      "learning_rate": 9.373851791988839e-07,
-      "long_answer_loss": 0.0762,
-      "loss": 0.0864,
-      "short_answer_loss": NaN,
-      "step": 1082,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0854,
-      "grad_norm": 1.40625,
-      "learning_rate": 9.249186665943072e-07,
-      "long_answer_loss": 0.0854,
-      "loss": 0.0838,
-      "short_answer_loss": NaN,
-      "step": 1083,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.087,
-      "grad_norm": 1.4453125,
-      "learning_rate": 9.125324220297243e-07,
-      "long_answer_loss": 0.087,
-      "loss": 0.0867,
-      "short_answer_loss": NaN,
-      "step": 1084,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0991,
-      "grad_norm": 1.359375,
-      "learning_rate": 9.002265313982336e-07,
-      "long_answer_loss": 0.0991,
-      "loss": 0.0804,
-      "short_answer_loss": NaN,
-      "step": 1085,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0868,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.880010800357053e-07,
-      "long_answer_loss": 0.0868,
-      "loss": 0.0905,
-      "short_answer_loss": NaN,
-      "step": 1086,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0932,
-      "grad_norm": 1.3515625,
-      "learning_rate": 8.758561527202169e-07,
-      "long_answer_loss": 0.0932,
-      "loss": 0.0892,
-      "short_answer_loss": NaN,
-      "step": 1087,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0954,
-      "grad_norm": 1.40625,
-      "learning_rate": 8.637918336714291e-07,
-      "long_answer_loss": 0.0954,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 1088,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0752,
-      "grad_norm": 1.3359375,
-      "learning_rate": 8.518082065500296e-07,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0774,
-      "short_answer_loss": NaN,
-      "step": 1089,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0756,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.399053544571434e-07,
-      "long_answer_loss": 0.0756,
-      "loss": 0.0909,
-      "short_answer_loss": NaN,
-      "step": 1090,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0844,
-      "grad_norm": 1.359375,
-      "learning_rate": 8.280833599337512e-07,
-      "long_answer_loss": 0.0844,
-      "loss": 0.0809,
-      "short_answer_loss": NaN,
-      "step": 1091,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.1157,
-      "grad_norm": 1.4296875,
-      "learning_rate": 8.16342304960127e-07,
-      "long_answer_loss": 0.1157,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 1092,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0911,
-      "grad_norm": 1.4609375,
-      "learning_rate": 8.046822709552704e-07,
-      "long_answer_loss": 0.0911,
-      "loss": 0.0874,
-      "short_answer_loss": NaN,
-      "step": 1093,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.076,
-      "grad_norm": 1.375,
-      "learning_rate": 7.931033387763229e-07,
-      "long_answer_loss": 0.076,
-      "loss": 0.0838,
-      "short_answer_loss": NaN,
-      "step": 1094,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0807,
-      "grad_norm": 1.390625,
-      "learning_rate": 7.816055887180318e-07,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0736,
-      "short_answer_loss": NaN,
-      "step": 1095,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.1084,
-      "grad_norm": 1.3984375,
-      "learning_rate": 7.701891005121794e-07,
-      "long_answer_loss": 0.1084,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 1096,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0891,
-      "grad_norm": 1.40625,
-      "learning_rate": 7.588539533270295e-07,
-      "long_answer_loss": 0.0891,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 1097,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0747,
-      "grad_norm": 1.3828125,
-      "learning_rate": 7.476002257667897e-07,
-      "long_answer_loss": 0.0747,
-      "loss": 0.0752,
-      "short_answer_loss": NaN,
-      "step": 1098,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0709,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.364279958710516e-07,
-      "long_answer_loss": 0.0709,
-      "loss": 0.0832,
-      "short_answer_loss": NaN,
-      "step": 1099,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0963,
-      "grad_norm": 1.359375,
-      "learning_rate": 7.2533734111426e-07,
-      "long_answer_loss": 0.0963,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 1100,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0834,
-      "grad_norm": 1.5390625,
-      "learning_rate": 7.143283384051763e-07,
-      "long_answer_loss": 0.0834,
-      "loss": 0.087,
-      "short_answer_loss": NaN,
-      "step": 1101,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0738,
-      "grad_norm": 1.3046875,
-      "learning_rate": 7.034010640863301e-07,
-      "long_answer_loss": 0.0738,
-      "loss": 0.077,
-      "short_answer_loss": NaN,
-      "step": 1102,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0679,
-      "grad_norm": 1.4609375,
-      "learning_rate": 6.925555939335124e-07,
-      "long_answer_loss": 0.0679,
-      "loss": 0.0847,
-      "short_answer_loss": NaN,
-      "step": 1103,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0554,
-      "grad_norm": 1.4375,
-      "learning_rate": 6.81792003155235e-07,
-      "long_answer_loss": 0.0554,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 1104,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0686,
-      "grad_norm": 1.4296875,
-      "learning_rate": 6.711103663922094e-07,
-      "long_answer_loss": 0.0686,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 1105,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0825,
-      "grad_norm": 1.34375,
-      "learning_rate": 6.605107577168341e-07,
-      "long_answer_loss": 0.0825,
-      "loss": 0.0847,
-      "short_answer_loss": NaN,
-      "step": 1106,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0875,
-      "grad_norm": 1.4296875,
-      "learning_rate": 6.499932506326828e-07,
-      "long_answer_loss": 0.0875,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 1107,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0868,
-      "grad_norm": 1.3359375,
-      "learning_rate": 6.395579180739866e-07,
-      "long_answer_loss": 0.0868,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 1108,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.1108,
-      "grad_norm": 1.40625,
-      "learning_rate": 6.292048324051378e-07,
-      "long_answer_loss": 0.1108,
-      "loss": 0.0874,
-      "short_answer_loss": NaN,
-      "step": 1109,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0865,
-      "grad_norm": 1.4609375,
-      "learning_rate": 6.189340654201751e-07,
-      "long_answer_loss": 0.0865,
-      "loss": 0.0787,
-      "short_answer_loss": NaN,
-      "step": 1110,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0636,
-      "grad_norm": 1.3984375,
-      "learning_rate": 6.08745688342302e-07,
-      "long_answer_loss": 0.0636,
-      "loss": 0.0784,
-      "short_answer_loss": NaN,
-      "step": 1111,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0621,
-      "grad_norm": 1.3828125,
-      "learning_rate": 5.986397718233838e-07,
-      "long_answer_loss": 0.0621,
-      "loss": 0.0949,
-      "short_answer_loss": NaN,
-      "step": 1112,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0806,
-      "grad_norm": 1.3671875,
-      "learning_rate": 5.886163859434482e-07,
-      "long_answer_loss": 0.0806,
-      "loss": 0.0812,
-      "short_answer_loss": NaN,
-      "step": 1113,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0829,
-      "grad_norm": 1.4453125,
-      "learning_rate": 5.786756002102239e-07,
-      "long_answer_loss": 0.0829,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 1114,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0766,
-      "grad_norm": 1.484375,
-      "learning_rate": 5.688174835586382e-07,
-      "long_answer_loss": 0.0766,
-      "loss": 0.0884,
-      "short_answer_loss": NaN,
-      "step": 1115,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0892,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.590421043503422e-07,
-      "long_answer_loss": 0.0892,
-      "loss": 0.085,
-      "short_answer_loss": NaN,
-      "step": 1116,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0944,
-      "grad_norm": 1.453125,
-      "learning_rate": 5.493495303732415e-07,
-      "long_answer_loss": 0.0944,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 1117,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0704,
-      "grad_norm": 1.421875,
-      "learning_rate": 5.397398288410293e-07,
-      "long_answer_loss": 0.0704,
-      "loss": 0.0751,
-      "short_answer_loss": NaN,
-      "step": 1118,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0949,
-      "grad_norm": 1.3359375,
-      "learning_rate": 5.30213066392704e-07,
-      "long_answer_loss": 0.0949,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 1119,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0618,
-      "grad_norm": 1.484375,
-      "learning_rate": 5.207693090921325e-07,
-      "long_answer_loss": 0.0618,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 1120,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0886,
-      "grad_norm": 1.375,
-      "learning_rate": 5.114086224275671e-07,
-      "long_answer_loss": 0.0886,
-      "loss": 0.0895,
-      "short_answer_loss": NaN,
-      "step": 1121,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.1024,
-      "grad_norm": 1.5078125,
-      "learning_rate": 5.021310713112057e-07,
-      "long_answer_loss": 0.1024,
-      "loss": 0.0876,
-      "short_answer_loss": NaN,
-      "step": 1122,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0832,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.929367200787405e-07,
-      "long_answer_loss": 0.0832,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 1123,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0724,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.838256324889046e-07,
-      "long_answer_loss": 0.0724,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 1124,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0846,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.7479787172303727e-07,
-      "long_answer_loss": 0.0846,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 1125,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0958,
-      "grad_norm": 1.34375,
-      "learning_rate": 4.6585350038464565e-07,
-      "long_answer_loss": 0.0958,
-      "loss": 0.0832,
-      "short_answer_loss": NaN,
-      "step": 1126,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0911,
-      "grad_norm": 1.328125,
-      "learning_rate": 4.569925804989647e-07,
-      "long_answer_loss": 0.0911,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 1127,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0884,
-      "grad_norm": 1.46875,
-      "learning_rate": 4.4821517351253533e-07,
-      "long_answer_loss": 0.0884,
-      "loss": 0.0843,
-      "short_answer_loss": NaN,
-      "step": 1128,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0768,
-      "grad_norm": 1.40625,
-      "learning_rate": 4.3952134029276587e-07,
-      "long_answer_loss": 0.0768,
-      "loss": 0.0819,
-      "short_answer_loss": NaN,
-      "step": 1129,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0706,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.309111411275241e-07,
-      "long_answer_loss": 0.0706,
-      "loss": 0.0904,
-      "short_answer_loss": NaN,
-      "step": 1130,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.1012,
-      "grad_norm": 1.3671875,
-      "learning_rate": 4.223846357247124e-07,
-      "long_answer_loss": 0.1012,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 1131,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0776,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.139418832118505e-07,
-      "long_answer_loss": 0.0776,
-      "loss": 0.081,
-      "short_answer_loss": NaN,
-      "step": 1132,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0971,
-      "grad_norm": 1.6328125,
-      "learning_rate": 4.0558294213567105e-07,
-      "long_answer_loss": 0.0971,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 1133,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0809,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.973078704617175e-07,
-      "long_answer_loss": 0.0809,
-      "loss": 0.0823,
-      "short_answer_loss": NaN,
-      "step": 1134,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0923,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.89116725573925e-07,
-      "long_answer_loss": 0.0923,
-      "loss": 0.0855,
-      "short_answer_loss": NaN,
-      "step": 1135,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0775,
-      "grad_norm": 1.328125,
-      "learning_rate": 3.810095642742414e-07,
-      "long_answer_loss": 0.0775,
-      "loss": 0.0767,
-      "short_answer_loss": NaN,
-      "step": 1136,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0858,
-      "grad_norm": 1.3671875,
-      "learning_rate": 3.7298644278222494e-07,
-      "long_answer_loss": 0.0858,
-      "loss": 0.0772,
-      "short_answer_loss": NaN,
-      "step": 1137,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0852,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.6504741673465264e-07,
-      "long_answer_loss": 0.0852,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 1138,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0737,
-      "grad_norm": 1.375,
-      "learning_rate": 3.57192541185139e-07,
-      "long_answer_loss": 0.0737,
-      "loss": 0.0775,
-      "short_answer_loss": NaN,
-      "step": 1139,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.087,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.4942187060375277e-07,
-      "long_answer_loss": 0.087,
-      "loss": 0.0792,
-      "short_answer_loss": NaN,
-      "step": 1140,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0693,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.417354588766353e-07,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0809,
-      "short_answer_loss": NaN,
-      "step": 1141,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0885,
-      "grad_norm": 1.5078125,
-      "learning_rate": 3.34133359305637e-07,
-      "long_answer_loss": 0.0885,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 1142,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0765,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.266156246079316e-07,
-      "long_answer_loss": 0.0765,
-      "loss": 0.0848,
-      "short_answer_loss": NaN,
-      "step": 1143,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0744,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.1918230691566906e-07,
-      "long_answer_loss": 0.0744,
-      "loss": 0.0893,
-      "short_answer_loss": NaN,
-      "step": 1144,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0869,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.1183345777559964e-07,
-      "long_answer_loss": 0.0869,
-      "loss": 0.0794,
-      "short_answer_loss": NaN,
-      "step": 1145,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0858,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.045691281487198e-07,
-      "long_answer_loss": 0.0858,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 1146,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.104,
-      "grad_norm": 1.484375,
-      "learning_rate": 2.9738936840992967e-07,
-      "long_answer_loss": 0.104,
-      "loss": 0.0841,
-      "short_answer_loss": NaN,
-      "step": 1147,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0851,
-      "grad_norm": 1.3515625,
-      "learning_rate": 2.9029422834766645e-07,
-      "long_answer_loss": 0.0851,
-      "loss": 0.0788,
-      "short_answer_loss": NaN,
-      "step": 1148,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.081,
-      "grad_norm": 1.390625,
-      "learning_rate": 2.832837571635688e-07,
-      "long_answer_loss": 0.081,
-      "loss": 0.0788,
-      "short_answer_loss": NaN,
-      "step": 1149,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.08,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.763580034721394e-07,
-      "long_answer_loss": 0.08,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 1150,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.081,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.6951701530039676e-07,
-      "long_answer_loss": 0.081,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 1151,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0649,
-      "grad_norm": 1.515625,
-      "learning_rate": 2.627608400875503e-07,
-      "long_answer_loss": 0.0649,
-      "loss": 0.0764,
-      "short_answer_loss": NaN,
-      "step": 1152,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0826,
-      "grad_norm": 1.5703125,
-      "learning_rate": 2.5608952468467175e-07,
-      "long_answer_loss": 0.0826,
-      "loss": 0.0866,
-      "short_answer_loss": NaN,
-      "step": 1153,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0923,
-      "grad_norm": 1.328125,
-      "learning_rate": 2.495031153543631e-07,
-      "long_answer_loss": 0.0923,
-      "loss": 0.0838,
-      "short_answer_loss": NaN,
-      "step": 1154,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0851,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.430016577704461e-07,
-      "long_answer_loss": 0.0851,
-      "loss": 0.0851,
-      "short_answer_loss": NaN,
-      "step": 1155,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0785,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.365851970176358e-07,
-      "long_answer_loss": 0.0785,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 1156,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0705,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.3025377759123279e-07,
-      "long_answer_loss": 0.0705,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 1157,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0859,
-      "grad_norm": 1.390625,
-      "learning_rate": 2.240074433968134e-07,
-      "long_answer_loss": 0.0859,
-      "loss": 0.0822,
-      "short_answer_loss": NaN,
-      "step": 1158,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0678,
-      "grad_norm": 1.3125,
-      "learning_rate": 2.178462377499302e-07,
-      "long_answer_loss": 0.0678,
-      "loss": 0.0768,
-      "short_answer_loss": NaN,
-      "step": 1159,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.1107,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.1177020337579818e-07,
-      "long_answer_loss": 0.1107,
-      "loss": 0.0831,
-      "short_answer_loss": NaN,
-      "step": 1160,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.09,
-      "grad_norm": 1.5078125,
-      "learning_rate": 2.0577938240901873e-07,
-      "long_answer_loss": 0.09,
-      "loss": 0.0836,
-      "short_answer_loss": NaN,
-      "step": 1161,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.1141,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.998738163932659e-07,
-      "long_answer_loss": 0.1141,
-      "loss": 0.0827,
-      "short_answer_loss": NaN,
-      "step": 1162,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0861,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.9405354628101448e-07,
-      "long_answer_loss": 0.0861,
-      "loss": 0.0859,
-      "short_answer_loss": NaN,
-      "step": 1163,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0883,
-      "grad_norm": 1.375,
-      "learning_rate": 1.8831861243324978e-07,
-      "long_answer_loss": 0.0883,
-      "loss": 0.0789,
-      "short_answer_loss": NaN,
-      "step": 1164,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0847,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.8266905461918475e-07,
-      "long_answer_loss": 0.0847,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 1165,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0852,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.7710491201599065e-07,
-      "long_answer_loss": 0.0852,
-      "loss": 0.08,
-      "short_answer_loss": NaN,
-      "step": 1166,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0977,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.7162622320852084e-07,
-      "long_answer_loss": 0.0977,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 1167,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0761,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.6623302618904713e-07,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 1168,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0999,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.6092535835699058e-07,
-      "long_answer_loss": 0.0999,
-      "loss": 0.0907,
-      "short_answer_loss": NaN,
-      "step": 1169,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0967,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.5570325651866475e-07,
-      "long_answer_loss": 0.0967,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 1170,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0865,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.5056675688702449e-07,
-      "long_answer_loss": 0.0865,
-      "loss": 0.0782,
-      "short_answer_loss": NaN,
-      "step": 1171,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0635,
-      "grad_norm": 1.375,
-      "learning_rate": 1.4551589508141062e-07,
-      "long_answer_loss": 0.0635,
-      "loss": 0.0797,
-      "short_answer_loss": NaN,
-      "step": 1172,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0787,
-      "grad_norm": 1.359375,
-      "learning_rate": 1.405507061273001e-07,
-      "long_answer_loss": 0.0787,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 1173,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0891,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.3567122445607427e-07,
-      "long_answer_loss": 0.0891,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 1174,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.1124,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.3087748390476356e-07,
-      "long_answer_loss": 0.1124,
-      "loss": 0.0879,
-      "short_answer_loss": NaN,
-      "step": 1175,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0836,
-      "grad_norm": 1.3203125,
-      "learning_rate": 1.2616951771582674e-07,
-      "long_answer_loss": 0.0836,
-      "loss": 0.0808,
-      "short_answer_loss": NaN,
-      "step": 1176,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0714,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.2154735853691923e-07,
-      "long_answer_loss": 0.0714,
-      "loss": 0.0775,
-      "short_answer_loss": NaN,
-      "step": 1177,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0685,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.1701103842065298e-07,
-      "long_answer_loss": 0.0685,
-      "loss": 0.0837,
-      "short_answer_loss": NaN,
-      "step": 1178,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0862,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.1256058882439386e-07,
-      "long_answer_loss": 0.0862,
-      "loss": 0.0776,
-      "short_answer_loss": NaN,
-      "step": 1179,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0845,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.0819604061002852e-07,
-      "long_answer_loss": 0.0845,
-      "loss": 0.0796,
-      "short_answer_loss": NaN,
-      "step": 1180,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.1098,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.0391742404375765e-07,
-      "long_answer_loss": 0.1098,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 1181,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.1155,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.972476879588494e-08,
-      "long_answer_loss": 0.1155,
-      "loss": 0.081,
-      "short_answer_loss": NaN,
-      "step": 1182,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.1045,
-      "grad_norm": 1.484375,
-      "learning_rate": 9.5618103940609e-08,
-      "long_answer_loss": 0.1045,
-      "loss": 0.0891,
-      "short_answer_loss": NaN,
-      "step": 1183,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0814,
-      "grad_norm": 1.40625,
-      "learning_rate": 9.159745795582209e-08,
-      "long_answer_loss": 0.0814,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 1184,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0752,
-      "grad_norm": 1.3203125,
-      "learning_rate": 8.766285872291725e-08,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0789,
-      "short_answer_loss": NaN,
-      "step": 1185,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.073,
-      "grad_norm": 1.3359375,
-      "learning_rate": 8.38143335265898e-08,
-      "long_answer_loss": 0.073,
-      "loss": 0.0779,
-      "short_answer_loss": NaN,
-      "step": 1186,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0729,
-      "grad_norm": 1.5,
-      "learning_rate": 8.005190905464866e-08,
-      "long_answer_loss": 0.0729,
-      "loss": 0.0849,
-      "short_answer_loss": NaN,
-      "step": 1187,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0817,
-      "grad_norm": 1.390625,
-      "learning_rate": 7.637561139783589e-08,
-      "long_answer_loss": 0.0817,
-      "loss": 0.0807,
-      "short_answer_loss": NaN,
-      "step": 1188,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0808,
-      "grad_norm": 1.4609375,
-      "learning_rate": 7.278546604963937e-08,
-      "long_answer_loss": 0.0808,
-      "loss": 0.0888,
-      "short_answer_loss": NaN,
-      "step": 1189,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0643,
-      "grad_norm": 1.40625,
-      "learning_rate": 6.928149790611932e-08,
-      "long_answer_loss": 0.0643,
-      "loss": 0.0778,
-      "short_answer_loss": NaN,
-      "step": 1190,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.077,
-      "grad_norm": 1.3828125,
-      "learning_rate": 6.586373126573759e-08,
-      "long_answer_loss": 0.077,
-      "loss": 0.0801,
-      "short_answer_loss": NaN,
-      "step": 1191,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0875,
-      "grad_norm": 1.3671875,
-      "learning_rate": 6.253218982918418e-08,
-      "long_answer_loss": 0.0875,
-      "loss": 0.086,
-      "short_answer_loss": NaN,
-      "step": 1192,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.07,
-      "grad_norm": 1.3359375,
-      "learning_rate": 5.928689669921772e-08,
-      "long_answer_loss": 0.07,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 1193,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0873,
-      "grad_norm": 1.3203125,
-      "learning_rate": 5.612787438050299e-08,
-      "long_answer_loss": 0.0873,
-      "loss": 0.0777,
-      "short_answer_loss": NaN,
-      "step": 1194,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.1155,
-      "grad_norm": 1.3828125,
-      "learning_rate": 5.305514477945278e-08,
-      "long_answer_loss": 0.1155,
-      "loss": 0.0843,
-      "short_answer_loss": NaN,
-      "step": 1195,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0967,
-      "grad_norm": 1.65625,
-      "learning_rate": 5.006872920408079e-08,
-      "long_answer_loss": 0.0967,
-      "loss": 0.0931,
-      "short_answer_loss": NaN,
-      "step": 1196,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0776,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.716864836385171e-08,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0784,
-      "short_answer_loss": NaN,
-      "step": 1197,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0943,
-      "grad_norm": 1.5234375,
-      "learning_rate": 4.435492236953415e-08,
-      "long_answer_loss": 0.0943,
-      "loss": 0.0904,
-      "short_answer_loss": NaN,
-      "step": 1198,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0857,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.1627570733067386e-08,
-      "long_answer_loss": 0.0857,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 1199,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0795,
-      "grad_norm": 1.359375,
-      "learning_rate": 3.898661236742124e-08,
-      "long_answer_loss": 0.0795,
-      "loss": 0.0795,
-      "short_answer_loss": NaN,
-      "step": 1200,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.064,
-      "grad_norm": 1.3515625,
-      "learning_rate": 3.643206558646833e-08,
-      "long_answer_loss": 0.064,
-      "loss": 0.0757,
-      "short_answer_loss": NaN,
-      "step": 1201,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0763,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.39639481048537e-08,
-      "long_answer_loss": 0.0763,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 1202,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0784,
-      "grad_norm": 1.3671875,
-      "learning_rate": 3.158227703787264e-08,
-      "long_answer_loss": 0.0784,
-      "loss": 0.0806,
-      "short_answer_loss": NaN,
-      "step": 1203,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0618,
-      "grad_norm": 1.5,
-      "learning_rate": 2.9287068901356907e-08,
-      "long_answer_loss": 0.0618,
-      "loss": 0.0823,
-      "short_answer_loss": NaN,
-      "step": 1204,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0754,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.7078339611552595e-08,
-      "long_answer_loss": 0.0754,
-      "loss": 0.0846,
-      "short_answer_loss": NaN,
-      "step": 1205,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0835,
-      "grad_norm": 1.4921875,
-      "learning_rate": 2.4956104485014674e-08,
-      "long_answer_loss": 0.0835,
-      "loss": 0.0816,
-      "short_answer_loss": NaN,
-      "step": 1206,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0722,
-      "grad_norm": 1.328125,
-      "learning_rate": 2.292037823849874e-08,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0784,
-      "short_answer_loss": NaN,
-      "step": 1207,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.1098,
-      "grad_norm": 1.3671875,
-      "learning_rate": 2.0971174988863862e-08,
-      "long_answer_loss": 0.1098,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 1208,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0786,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.910850825296573e-08,
-      "long_answer_loss": 0.0786,
-      "loss": 0.0785,
-      "short_answer_loss": NaN,
-      "step": 1209,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.093,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.7332390947567835e-08,
-      "long_answer_loss": 0.093,
-      "loss": 0.0845,
-      "short_answer_loss": NaN,
-      "step": 1210,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0732,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.5642835389256817e-08,
-      "long_answer_loss": 0.0732,
-      "loss": 0.0809,
-      "short_answer_loss": NaN,
-      "step": 1211,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0713,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.4039853294346705e-08,
-      "long_answer_loss": 0.0713,
-      "loss": 0.0842,
-      "short_answer_loss": NaN,
-      "step": 1212,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0793,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2523455778806758e-08,
-      "long_answer_loss": 0.0793,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 1213,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0905,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.1093653358176804e-08,
-      "long_answer_loss": 0.0905,
-      "loss": 0.0784,
-      "short_answer_loss": NaN,
-      "step": 1214,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0979,
-      "grad_norm": 1.421875,
-      "learning_rate": 9.7504559475034e-09,
-      "long_answer_loss": 0.0979,
-      "loss": 0.0912,
-      "short_answer_loss": NaN,
-      "step": 1215,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0766,
-      "grad_norm": 1.359375,
-      "learning_rate": 8.493872861260744e-09,
-      "long_answer_loss": 0.0766,
-      "loss": 0.0797,
-      "short_answer_loss": NaN,
-      "step": 1216,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0754,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.323912813295142e-09,
-      "long_answer_loss": 0.0754,
-      "loss": 0.0914,
-      "short_answer_loss": NaN,
-      "step": 1217,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0735,
-      "grad_norm": 1.421875,
-      "learning_rate": 6.240583916758414e-09,
-      "long_answer_loss": 0.0735,
-      "loss": 0.0827,
-      "short_answer_loss": NaN,
-      "step": 1218,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0803,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.2438936840551454e-09,
-      "long_answer_loss": 0.0803,
-      "loss": 0.0869,
-      "short_answer_loss": NaN,
-      "step": 1219,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0725,
-      "grad_norm": 1.3515625,
-      "learning_rate": 4.3338490267871845e-09,
-      "long_answer_loss": 0.0725,
-      "loss": 0.0842,
-      "short_answer_loss": NaN,
-      "step": 1220,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0937,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.5104562557120026e-09,
-      "long_answer_loss": 0.0937,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 1221,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0521,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.7737210806899618e-09,
-      "long_answer_loss": 0.0521,
-      "loss": 0.0773,
-      "short_answer_loss": NaN,
-      "step": 1222,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0802,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.1236486106523957e-09,
-      "long_answer_loss": 0.0802,
-      "loss": 0.0854,
-      "short_answer_loss": NaN,
-      "step": 1223,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0651,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.5602433535627515e-09,
-      "long_answer_loss": 0.0651,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 1224,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0774,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.0835092163860582e-09,
-      "long_answer_loss": 0.0774,
-      "loss": 0.0771,
-      "short_answer_loss": NaN,
-      "step": 1225,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.095,
-      "grad_norm": 1.46875,
-      "learning_rate": 6.934495050611723e-10,
-      "long_answer_loss": 0.095,
-      "loss": 0.0827,
-      "short_answer_loss": NaN,
-      "step": 1226,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0651,
-      "grad_norm": 1.34375,
-      "learning_rate": 3.9006692447857153e-10,
-      "long_answer_loss": 0.0651,
-      "loss": 0.0788,
-      "short_answer_loss": NaN,
-      "step": 1227,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0897,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.7336357846231556e-10,
-      "long_answer_loss": 0.0897,
-      "loss": 0.0797,
-      "short_answer_loss": NaN,
-      "step": 1228,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "full_loss": 0.0987,
-      "grad_norm": 1.3515625,
-      "learning_rate": 4.3340969753391526e-11,
-      "long_answer_loss": 0.0987,
-      "loss": 0.0815,
-      "short_answer_loss": NaN,
-      "step": 1229,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "full_loss": 0.0642,
-      "grad_norm": 1.421875,
-      "learning_rate": 0.0,
-      "long_answer_loss": 0.0642,
-      "loss": 0.0863,
-      "short_answer_loss": NaN,
-      "step": 1230,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "step": 1230,
-      "total_flos": 9.273133451978998e+17,
-      "train_loss": 0.14522774513295997,
-      "train_runtime": 5219.7207,
-      "train_samples_per_second": 30.187,
-      "train_steps_per_second": 0.236
-    }
-  ],
-  "logging_steps": 1.0,
-  "max_steps": 1230,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
-  "save_steps": 1.0,
-  "total_flos": 9.273133451978998e+17,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}