diff --git "a/trainer_state.json" "b/trainer_state.json"
deleted file mode 100755--- "a/trainer_state.json"
+++ /dev/null
@@ -1,13626 +0,0 @@
-{
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 1.998383185125303,
-  "eval_steps": 25.0,
-  "global_step": 1236,
-  "is_hyper_param_search": false,
-  "is_local_process_zero": true,
-  "is_world_process_zero": true,
-  "log_history": [
-    {
-      "epoch": 0.0,
-      "full_loss": 0.5913,
-      "grad_norm": 11.4375,
-      "learning_rate": 6.578947368421053e-07,
-      "long_answer_loss": 0.5913,
-      "loss": 0.5772,
-      "short_answer_loss": NaN,
-      "step": 1,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.0,
-      "full_loss": 0.5407,
-      "grad_norm": 11.6875,
-      "learning_rate": 1.3157894736842106e-06,
-      "long_answer_loss": 0.5407,
-      "loss": 0.5984,
-      "short_answer_loss": NaN,
-      "step": 2,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.0,
-      "full_loss": 0.5632,
-      "grad_norm": 11.0,
-      "learning_rate": 1.9736842105263157e-06,
-      "long_answer_loss": 0.5632,
-      "loss": 0.5768,
-      "short_answer_loss": NaN,
-      "step": 3,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.5517,
-      "grad_norm": 10.5,
-      "learning_rate": 2.631578947368421e-06,
-      "long_answer_loss": 0.5517,
-      "loss": 0.5593,
-      "short_answer_loss": NaN,
-      "step": 4,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.4862,
-      "grad_norm": 9.0625,
-      "learning_rate": 3.2894736842105265e-06,
-      "long_answer_loss": 0.4862,
-      "loss": 0.5092,
-      "short_answer_loss": NaN,
-      "step": 5,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.501,
-      "grad_norm": 8.25,
-      "learning_rate": 3.9473684210526315e-06,
-      "long_answer_loss": 0.501,
-      "loss": 0.489,
-      "short_answer_loss": NaN,
-      "step": 6,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.4114,
-      "grad_norm": 7.0,
-      "learning_rate": 4.605263157894737e-06,
-      "long_answer_loss": 0.4114,
-      "loss": 0.4271,
-      "short_answer_loss": NaN,
-      "step": 7,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.357,
-      "grad_norm": 5.3125,
-      "learning_rate": 5.263157894736842e-06,
-      "long_answer_loss": 0.357,
-      "loss": 0.3714,
-      "short_answer_loss": NaN,
-      "step": 8,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.01,
-      "full_loss": 0.3179,
-      "grad_norm": 6.6875,
-      "learning_rate": 5.921052631578948e-06,
-      "long_answer_loss": 0.3179,
-      "loss": 0.3373,
-      "short_answer_loss": NaN,
-      "step": 9,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2982,
-      "grad_norm": 6.3125,
-      "learning_rate": 6.578947368421053e-06,
-      "long_answer_loss": 0.2982,
-      "loss": 0.3161,
-      "short_answer_loss": NaN,
-      "step": 10,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2744,
-      "grad_norm": 4.96875,
-      "learning_rate": 7.236842105263158e-06,
-      "long_answer_loss": 0.2744,
-      "loss": 0.2775,
-      "short_answer_loss": NaN,
-      "step": 11,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2795,
-      "grad_norm": 4.21875,
-      "learning_rate": 7.894736842105263e-06,
-      "long_answer_loss": 0.2795,
-      "loss": 0.2757,
-      "short_answer_loss": NaN,
-      "step": 12,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2391,
-      "grad_norm": 3.515625,
-      "learning_rate": 8.552631578947368e-06,
-      "long_answer_loss": 0.2391,
-      "loss": 0.241,
-      "short_answer_loss": NaN,
-      "step": 13,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2621,
-      "grad_norm": 3.0625,
-      "learning_rate": 9.210526315789474e-06,
-      "long_answer_loss": 0.2621,
-      "loss": 0.2527,
-      "short_answer_loss": NaN,
-      "step": 14,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.02,
-      "full_loss": 0.2162,
-      "grad_norm": 2.875,
-      "learning_rate": 9.868421052631579e-06,
-      "long_answer_loss": 0.2162,
-      "loss": 0.2324,
-      "short_answer_loss": NaN,
-      "step": 15,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2265,
-      "grad_norm": 3.078125,
-      "learning_rate": 1.0526315789473684e-05,
-      "long_answer_loss": 0.2265,
-      "loss": 0.238,
-      "short_answer_loss": NaN,
-      "step": 16,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2399,
-      "grad_norm": 2.96875,
-      "learning_rate": 1.118421052631579e-05,
-      "long_answer_loss": 0.2399,
-      "loss": 0.2288,
-      "short_answer_loss": NaN,
-      "step": 17,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.1865,
-      "grad_norm": 2.65625,
-      "learning_rate": 1.1842105263157895e-05,
-      "long_answer_loss": 0.1865,
-      "loss": 0.209,
-      "short_answer_loss": NaN,
-      "step": 18,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.1784,
-      "grad_norm": 2.703125,
-      "learning_rate": 1.25e-05,
-      "long_answer_loss": 0.1784,
-      "loss": 0.2255,
-      "short_answer_loss": NaN,
-      "step": 19,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.2396,
-      "grad_norm": 2.859375,
-      "learning_rate": 1.3157894736842106e-05,
-      "long_answer_loss": 0.2396,
-      "loss": 0.2175,
-      "short_answer_loss": NaN,
-      "step": 20,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.03,
-      "full_loss": 0.1953,
-      "grad_norm": 2.796875,
-      "learning_rate": 1.3815789473684213e-05,
-      "long_answer_loss": 0.1953,
-      "loss": 0.2107,
-      "short_answer_loss": NaN,
-      "step": 21,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.1817,
-      "grad_norm": 2.625,
-      "learning_rate": 1.4473684210526317e-05,
-      "long_answer_loss": 0.1817,
-      "loss": 0.2208,
-      "short_answer_loss": NaN,
-      "step": 22,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.1913,
-      "grad_norm": 2.703125,
-      "learning_rate": 1.5131578947368422e-05,
-      "long_answer_loss": 0.1913,
-      "loss": 0.1938,
-      "short_answer_loss": NaN,
-      "step": 23,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.1947,
-      "grad_norm": 2.59375,
-      "learning_rate": 1.5789473684210526e-05,
-      "long_answer_loss": 0.1947,
-      "loss": 0.2017,
-      "short_answer_loss": NaN,
-      "step": 24,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2353,
-      "grad_norm": 2.59375,
-      "learning_rate": 1.6447368421052635e-05,
-      "long_answer_loss": 0.2353,
-      "loss": 0.2086,
-      "short_answer_loss": NaN,
-      "step": 25,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.2258,
-      "grad_norm": 2.578125,
-      "learning_rate": 1.7105263157894737e-05,
-      "long_answer_loss": 0.2258,
-      "loss": 0.1929,
-      "short_answer_loss": NaN,
-      "step": 26,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.04,
-      "full_loss": 0.1916,
-      "grad_norm": 2.8125,
-      "learning_rate": 1.7763157894736842e-05,
-      "long_answer_loss": 0.1916,
-      "loss": 0.1996,
-      "short_answer_loss": NaN,
-      "step": 27,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.185,
-      "grad_norm": 2.78125,
-      "learning_rate": 1.8421052631578947e-05,
-      "long_answer_loss": 0.185,
-      "loss": 0.193,
-      "short_answer_loss": NaN,
-      "step": 28,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.1788,
-      "grad_norm": 2.65625,
-      "learning_rate": 1.9078947368421056e-05,
-      "long_answer_loss": 0.1788,
-      "loss": 0.1975,
-      "short_answer_loss": NaN,
-      "step": 29,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.1976,
-      "grad_norm": 2.609375,
-      "learning_rate": 1.9736842105263158e-05,
-      "long_answer_loss": 0.1976,
-      "loss": 0.1802,
-      "short_answer_loss": NaN,
-      "step": 30,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.204,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.0394736842105264e-05,
-      "long_answer_loss": 0.204,
-      "loss": 0.1897,
-      "short_answer_loss": NaN,
-      "step": 31,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.191,
-      "grad_norm": 2.46875,
-      "learning_rate": 2.105263157894737e-05,
-      "long_answer_loss": 0.191,
-      "loss": 0.2029,
-      "short_answer_loss": NaN,
-      "step": 32,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.1831,
-      "grad_norm": 2.375,
-      "learning_rate": 2.1710526315789474e-05,
-      "long_answer_loss": 0.1831,
-      "loss": 0.1809,
-      "short_answer_loss": NaN,
-      "step": 33,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.05,
-      "full_loss": 0.2093,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.236842105263158e-05,
-      "long_answer_loss": 0.2093,
-      "loss": 0.193,
-      "short_answer_loss": NaN,
-      "step": 34,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2136,
-      "grad_norm": 2.640625,
-      "learning_rate": 2.3026315789473685e-05,
-      "long_answer_loss": 0.2136,
-      "loss": 0.1973,
-      "short_answer_loss": NaN,
-      "step": 35,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2073,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.368421052631579e-05,
-      "long_answer_loss": 0.2073,
-      "loss": 0.1839,
-      "short_answer_loss": NaN,
-      "step": 36,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.2317,
-      "grad_norm": 2.625,
-      "learning_rate": 2.4342105263157896e-05,
-      "long_answer_loss": 0.2317,
-      "loss": 0.1985,
-      "short_answer_loss": NaN,
-      "step": 37,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.1949,
-      "grad_norm": 2.578125,
-      "learning_rate": 2.5e-05,
-      "long_answer_loss": 0.1949,
-      "loss": 0.1837,
-      "short_answer_loss": NaN,
-      "step": 38,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.1792,
-      "grad_norm": 2.484375,
-      "learning_rate": 2.499995702005279e-05,
-      "long_answer_loss": 0.1792,
-      "loss": 0.1878,
-      "short_answer_loss": NaN,
-      "step": 39,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.06,
-      "full_loss": 0.1855,
-      "grad_norm": 2.25,
-      "learning_rate": 2.499982808050672e-05,
-      "long_answer_loss": 0.1855,
-      "loss": 0.1966,
-      "short_answer_loss": NaN,
-      "step": 40,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1857,
-      "grad_norm": 2.546875,
-      "learning_rate": 2.4999613182248482e-05,
-      "long_answer_loss": 0.1857,
-      "loss": 0.1934,
-      "short_answer_loss": NaN,
-      "step": 41,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1906,
-      "grad_norm": 2.4375,
-      "learning_rate": 2.499931232675589e-05,
-      "long_answer_loss": 0.1906,
-      "loss": 0.1937,
-      "short_answer_loss": NaN,
-      "step": 42,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1886,
-      "grad_norm": 2.515625,
-      "learning_rate": 2.499892551609786e-05,
-      "long_answer_loss": 0.1886,
-      "loss": 0.1995,
-      "short_answer_loss": NaN,
-      "step": 43,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1893,
-      "grad_norm": 2.734375,
-      "learning_rate": 2.499845275293441e-05,
-      "long_answer_loss": 0.1893,
-      "loss": 0.2026,
-      "short_answer_loss": NaN,
-      "step": 44,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.2057,
-      "grad_norm": 2.53125,
-      "learning_rate": 2.499789404051663e-05,
-      "long_answer_loss": 0.2057,
-      "loss": 0.2021,
-      "short_answer_loss": NaN,
-      "step": 45,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.07,
-      "full_loss": 0.1896,
-      "grad_norm": 2.5,
-      "learning_rate": 2.4997249382686673e-05,
-      "long_answer_loss": 0.1896,
-      "loss": 0.1881,
-      "short_answer_loss": NaN,
-      "step": 46,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.2066,
-      "grad_norm": 2.46875,
-      "learning_rate": 2.4996518783877716e-05,
-      "long_answer_loss": 0.2066,
-      "loss": 0.1984,
-      "short_answer_loss": NaN,
-      "step": 47,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1739,
-      "grad_norm": 2.515625,
-      "learning_rate": 2.4995702249113935e-05,
-      "long_answer_loss": 0.1739,
-      "loss": 0.1963,
-      "short_answer_loss": NaN,
-      "step": 48,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1877,
-      "grad_norm": 2.375,
-      "learning_rate": 2.499479978401047e-05,
-      "long_answer_loss": 0.1877,
-      "loss": 0.1921,
-      "short_answer_loss": NaN,
-      "step": 49,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1709,
-      "grad_norm": 2.375,
-      "learning_rate": 2.499381139477338e-05,
-      "long_answer_loss": 0.1709,
-      "loss": 0.1892,
-      "short_answer_loss": NaN,
-      "step": 50,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1591,
-      "grad_norm": 2.53125,
-      "learning_rate": 2.4992737088199623e-05,
-      "long_answer_loss": 0.1591,
-      "loss": 0.1913,
-      "short_answer_loss": NaN,
-      "step": 51,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.08,
-      "full_loss": 0.1878,
-      "grad_norm": 2.53125,
-      "learning_rate": 2.499157687167697e-05,
-      "long_answer_loss": 0.1878,
-      "loss": 0.1977,
-      "short_answer_loss": NaN,
-      "step": 52,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1668,
-      "grad_norm": 2.421875,
-      "learning_rate": 2.499033075318399e-05,
-      "long_answer_loss": 0.1668,
-      "loss": 0.1937,
-      "short_answer_loss": NaN,
-      "step": 53,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1727,
-      "grad_norm": 2.453125,
-      "learning_rate": 2.4988998741289986e-05,
-      "long_answer_loss": 0.1727,
-      "loss": 0.1807,
-      "short_answer_loss": NaN,
-      "step": 54,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1721,
-      "grad_norm": 2.9375,
-      "learning_rate": 2.4987580845154922e-05,
-      "long_answer_loss": 0.1721,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 55,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.2108,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.4986077074529374e-05,
-      "long_answer_loss": 0.2108,
-      "loss": 0.1986,
-      "short_answer_loss": NaN,
-      "step": 56,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1706,
-      "grad_norm": 2.453125,
-      "learning_rate": 2.498448743975446e-05,
-      "long_answer_loss": 0.1706,
-      "loss": 0.1781,
-      "short_answer_loss": NaN,
-      "step": 57,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.09,
-      "full_loss": 0.1635,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.498281195176177e-05,
-      "long_answer_loss": 0.1635,
-      "loss": 0.1883,
-      "short_answer_loss": NaN,
-      "step": 58,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.1799,
-      "grad_norm": 2.546875,
-      "learning_rate": 2.498105062207328e-05,
-      "long_answer_loss": 0.1799,
-      "loss": 0.1854,
-      "short_answer_loss": NaN,
-      "step": 59,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2089,
-      "grad_norm": 2.515625,
-      "learning_rate": 2.4979203462801287e-05,
-      "long_answer_loss": 0.2089,
-      "loss": 0.1871,
-      "short_answer_loss": NaN,
-      "step": 60,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.2018,
-      "grad_norm": 2.34375,
-      "learning_rate": 2.497727048664833e-05,
-      "long_answer_loss": 0.2018,
-      "loss": 0.1861,
-      "short_answer_loss": NaN,
-      "step": 61,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.203,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.497525170690707e-05,
-      "long_answer_loss": 0.203,
-      "loss": 0.1979,
-      "short_answer_loss": NaN,
-      "step": 62,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.1858,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.4973147137460246e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1876,
-      "short_answer_loss": NaN,
-      "step": 63,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.1,
-      "full_loss": 0.1554,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.4970956792780533e-05,
-      "long_answer_loss": 0.1554,
-      "loss": 0.1781,
-      "short_answer_loss": NaN,
-      "step": 64,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2165,
-      "grad_norm": 2.5,
-      "learning_rate": 2.4968680687930482e-05,
-      "long_answer_loss": 0.2165,
-      "loss": 0.1937,
-      "short_answer_loss": NaN,
-      "step": 65,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.186,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4966318838562392e-05,
-      "long_answer_loss": 0.186,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 66,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.1919,
-      "grad_norm": 2.34375,
-      "learning_rate": 2.49638712609182e-05,
-      "long_answer_loss": 0.1919,
-      "loss": 0.1886,
-      "short_answer_loss": NaN,
-      "step": 67,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.1953,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4961337971829397e-05,
-      "long_answer_loss": 0.1953,
-      "loss": 0.1902,
-      "short_answer_loss": NaN,
-      "step": 68,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.1847,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4958718988716885e-05,
-      "long_answer_loss": 0.1847,
-      "loss": 0.1858,
-      "short_answer_loss": NaN,
-      "step": 69,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.2127,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.4956014329590855e-05,
-      "long_answer_loss": 0.2127,
-      "loss": 0.1936,
-      "short_answer_loss": NaN,
-      "step": 70,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.11,
-      "full_loss": 0.1856,
-      "grad_norm": 2.421875,
-      "learning_rate": 2.495322401305069e-05,
-      "long_answer_loss": 0.1856,
-      "loss": 0.1982,
-      "short_answer_loss": NaN,
-      "step": 71,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.1997,
-      "grad_norm": 2.390625,
-      "learning_rate": 2.4950348058284813e-05,
-      "long_answer_loss": 0.1997,
-      "loss": 0.1893,
-      "short_answer_loss": NaN,
-      "step": 72,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2013,
-      "grad_norm": 2.125,
-      "learning_rate": 2.494738648507057e-05,
-      "long_answer_loss": 0.2013,
-      "loss": 0.1877,
-      "short_answer_loss": NaN,
-      "step": 73,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.1742,
-      "grad_norm": 2.1875,
-      "learning_rate": 2.494433931377408e-05,
-      "long_answer_loss": 0.1742,
-      "loss": 0.1841,
-      "short_answer_loss": NaN,
-      "step": 74,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2026,
-      "grad_norm": 2.21875,
-      "learning_rate": 2.4941206565350102e-05,
-      "long_answer_loss": 0.2026,
-      "loss": 0.1785,
-      "short_answer_loss": NaN,
-      "step": 75,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.2138,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.49379882613419e-05,
-      "long_answer_loss": 0.2138,
-      "loss": 0.1952,
-      "short_answer_loss": NaN,
-      "step": 76,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.12,
-      "full_loss": 0.183,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4934684423881074e-05,
-      "long_answer_loss": 0.183,
-      "loss": 0.1743,
-      "short_answer_loss": NaN,
-      "step": 77,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.1767,
-      "grad_norm": 2.390625,
-      "learning_rate": 2.4931295075687428e-05,
-      "long_answer_loss": 0.1767,
-      "loss": 0.1919,
-      "short_answer_loss": NaN,
-      "step": 78,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.1558,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4927820240068805e-05,
-      "long_answer_loss": 0.1558,
-      "loss": 0.1802,
-      "short_answer_loss": NaN,
-      "step": 79,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.1912,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.492425994092092e-05,
-      "long_answer_loss": 0.1912,
-      "loss": 0.1875,
-      "short_answer_loss": NaN,
-      "step": 80,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2034,
-      "grad_norm": 2.125,
-      "learning_rate": 2.4920614202727217e-05,
-      "long_answer_loss": 0.2034,
-      "loss": 0.1925,
-      "short_answer_loss": NaN,
-      "step": 81,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.1561,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4916883050558664e-05,
-      "long_answer_loss": 0.1561,
-      "loss": 0.1878,
-      "short_answer_loss": NaN,
-      "step": 82,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.13,
-      "full_loss": 0.2187,
-      "grad_norm": 2.34375,
-      "learning_rate": 2.491306651007363e-05,
-      "long_answer_loss": 0.2187,
-      "loss": 0.1993,
-      "short_answer_loss": NaN,
-      "step": 83,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.2146,
-      "grad_norm": 2.375,
-      "learning_rate": 2.490916460751766e-05,
-      "long_answer_loss": 0.2146,
-      "loss": 0.1946,
-      "short_answer_loss": NaN,
-      "step": 84,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1711,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4905177369723333e-05,
-      "long_answer_loss": 0.1711,
-      "loss": 0.1791,
-      "short_answer_loss": NaN,
-      "step": 85,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1823,
-      "grad_norm": 2.15625,
-      "learning_rate": 2.4901104824110042e-05,
-      "long_answer_loss": 0.1823,
-      "loss": 0.1865,
-      "short_answer_loss": NaN,
-      "step": 86,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1902,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.489694699868384e-05,
-      "long_answer_loss": 0.1902,
-      "loss": 0.1932,
-      "short_answer_loss": NaN,
-      "step": 87,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1712,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4892703922037225e-05,
-      "long_answer_loss": 0.1712,
-      "loss": 0.185,
-      "short_answer_loss": NaN,
-      "step": 88,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.14,
-      "full_loss": 0.1422,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.488837562334895e-05,
-      "long_answer_loss": 0.1422,
-      "loss": 0.1847,
-      "short_answer_loss": NaN,
-      "step": 89,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1879,
-      "grad_norm": 2.25,
-      "learning_rate": 2.4883962132383823e-05,
-      "long_answer_loss": 0.1879,
-      "loss": 0.19,
-      "short_answer_loss": NaN,
-      "step": 90,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2042,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4879463479492504e-05,
-      "long_answer_loss": 0.2042,
-      "loss": 0.1942,
-      "short_answer_loss": NaN,
-      "step": 91,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.2047,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.4874879695611287e-05,
-      "long_answer_loss": 0.2047,
-      "loss": 0.193,
-      "short_answer_loss": NaN,
-      "step": 92,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1753,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.4870210812261898e-05,
-      "long_answer_loss": 0.1753,
-      "loss": 0.1829,
-      "short_answer_loss": NaN,
-      "step": 93,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1802,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.486545686155128e-05,
-      "long_answer_loss": 0.1802,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 94,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.15,
-      "full_loss": 0.1443,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4860617876171355e-05,
-      "long_answer_loss": 0.1443,
-      "loss": 0.18,
-      "short_answer_loss": NaN,
-      "step": 95,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1755,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4855693889398822e-05,
-      "long_answer_loss": 0.1755,
-      "loss": 0.1902,
-      "short_answer_loss": NaN,
-      "step": 96,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1992,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.485068493509491e-05,
-      "long_answer_loss": 0.1992,
-      "loss": 0.1887,
-      "short_answer_loss": NaN,
-      "step": 97,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2405,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.4845591047705153e-05,
-      "long_answer_loss": 0.2405,
-      "loss": 0.1958,
-      "short_answer_loss": NaN,
-      "step": 98,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.2091,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.484041226225915e-05,
-      "long_answer_loss": 0.2091,
-      "loss": 0.1872,
-      "short_answer_loss": NaN,
-      "step": 99,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1757,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.4835148614370334e-05,
-      "long_answer_loss": 0.1757,
-      "loss": 0.1841,
-      "short_answer_loss": NaN,
-      "step": 100,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1818,
-      "grad_norm": 2.125,
-      "learning_rate": 2.482980014023571e-05,
-      "long_answer_loss": 0.1818,
-      "loss": 0.1852,
-      "short_answer_loss": NaN,
-      "step": 101,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.16,
-      "full_loss": 0.1899,
-      "grad_norm": 2.453125,
-      "learning_rate": 2.4824366876635623e-05,
-      "long_answer_loss": 0.1899,
-      "loss": 0.195,
-      "short_answer_loss": NaN,
-      "step": 102,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1858,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.481884886093349e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1946,
-      "short_answer_loss": NaN,
-      "step": 103,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1847,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.4813246131075564e-05,
-      "long_answer_loss": 0.1847,
-      "loss": 0.1895,
-      "short_answer_loss": NaN,
-      "step": 104,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1691,
-      "grad_norm": 2.1875,
-      "learning_rate": 2.480755872559064e-05,
-      "long_answer_loss": 0.1691,
-      "loss": 0.191,
-      "short_answer_loss": NaN,
-      "step": 105,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.2181,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.4801786683589824e-05,
-      "long_answer_loss": 0.2181,
-      "loss": 0.1879,
-      "short_answer_loss": NaN,
-      "step": 106,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1856,
-      "grad_norm": 2.25,
-      "learning_rate": 2.4795930044766247e-05,
-      "long_answer_loss": 0.1856,
-      "loss": 0.1899,
-      "short_answer_loss": NaN,
-      "step": 107,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.17,
-      "full_loss": 0.1754,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.4789988849394792e-05,
-      "long_answer_loss": 0.1754,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 108,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1835,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.478396313833182e-05,
-      "long_answer_loss": 0.1835,
-      "loss": 0.1983,
-      "short_answer_loss": NaN,
-      "step": 109,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.2057,
-      "grad_norm": 2.1875,
-      "learning_rate": 2.4777852953014896e-05,
-      "long_answer_loss": 0.2057,
-      "loss": 0.1931,
-      "short_answer_loss": NaN,
-      "step": 110,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1497,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.4771658335462483e-05,
-      "long_answer_loss": 0.1497,
-      "loss": 0.1912,
-      "short_answer_loss": NaN,
-      "step": 111,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1689,
-      "grad_norm": 2.21875,
-      "learning_rate": 2.476537932827368e-05,
-      "long_answer_loss": 0.1689,
-      "loss": 0.1817,
-      "short_answer_loss": NaN,
-      "step": 112,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.2024,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.4759015974627906e-05,
-      "long_answer_loss": 0.2024,
-      "loss": 0.1819,
-      "short_answer_loss": NaN,
-      "step": 113,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.18,
-      "full_loss": 0.1597,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.475256831828462e-05,
-      "long_answer_loss": 0.1597,
-      "loss": 0.1916,
-      "short_answer_loss": NaN,
-      "step": 114,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1845,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4746036403583012e-05,
-      "long_answer_loss": 0.1845,
-      "loss": 0.1897,
-      "short_answer_loss": NaN,
-      "step": 115,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1652,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.4739420275441694e-05,
-      "long_answer_loss": 0.1652,
-      "loss": 0.1878,
-      "short_answer_loss": NaN,
-      "step": 116,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1572,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.4732719979358403e-05,
-      "long_answer_loss": 0.1572,
-      "loss": 0.1766,
-      "short_answer_loss": NaN,
-      "step": 117,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1945,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.472593556140968e-05,
-      "long_answer_loss": 0.1945,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 118,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1754,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.4719067068250552e-05,
-      "long_answer_loss": 0.1754,
-      "loss": 0.1834,
-      "short_answer_loss": NaN,
-      "step": 119,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.19,
-      "full_loss": 0.1827,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4712114547114212e-05,
-      "long_answer_loss": 0.1827,
-      "loss": 0.1866,
-      "short_answer_loss": NaN,
-      "step": 120,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.1763,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.4705078045811704e-05,
-      "long_answer_loss": 0.1763,
-      "loss": 0.1771,
-      "short_answer_loss": NaN,
-      "step": 121,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.1739,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.469795761273157e-05,
-      "long_answer_loss": 0.1739,
-      "loss": 0.1826,
-      "short_answer_loss": NaN,
-      "step": 122,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.1831,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4690753296839558e-05,
-      "long_answer_loss": 0.1831,
-      "loss": 0.192,
-      "short_answer_loss": NaN,
-      "step": 123,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.1404,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4683465147678235e-05,
-      "long_answer_loss": 0.1404,
-      "loss": 0.1811,
-      "short_answer_loss": NaN,
-      "step": 124,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.1653,
-      "grad_norm": 2.140625,
-      "learning_rate": 2.4676093215366695e-05,
-      "long_answer_loss": 0.1653,
-      "loss": 0.185,
-      "short_answer_loss": NaN,
-      "step": 125,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.2,
-      "full_loss": 0.2176,
-      "grad_norm": 2.203125,
-      "learning_rate": 2.466863755060017e-05,
-      "long_answer_loss": 0.2176,
-      "loss": 0.1936,
-      "short_answer_loss": NaN,
-      "step": 126,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.2059,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.4661098204649717e-05,
-      "long_answer_loss": 0.2059,
-      "loss": 0.1855,
-      "short_answer_loss": NaN,
-      "step": 127,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1801,
-      "grad_norm": 2.40625,
-      "learning_rate": 2.4653475229361843e-05,
-      "long_answer_loss": 0.1801,
-      "loss": 0.193,
-      "short_answer_loss": NaN,
-      "step": 128,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1823,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.4645768677158165e-05,
-      "long_answer_loss": 0.1823,
-      "loss": 0.1789,
-      "short_answer_loss": NaN,
-      "step": 129,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1603,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.4637978601035033e-05,
-      "long_answer_loss": 0.1603,
-      "loss": 0.1858,
-      "short_answer_loss": NaN,
-      "step": 130,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1987,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.463010505456318e-05,
-      "long_answer_loss": 0.1987,
-      "loss": 0.1986,
-      "short_answer_loss": NaN,
-      "step": 131,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.21,
-      "full_loss": 0.1814,
-      "grad_norm": 2.171875,
-      "learning_rate": 2.4622148091887338e-05,
-      "long_answer_loss": 0.1814,
-      "loss": 0.1878,
-      "short_answer_loss": NaN,
-      "step": 132,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1871,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.4614107767725887e-05,
-      "long_answer_loss": 0.1871,
-      "loss": 0.1781,
-      "short_answer_loss": NaN,
-      "step": 133,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1763,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4605984137370452e-05,
-      "long_answer_loss": 0.1763,
-      "loss": 0.1949,
-      "short_answer_loss": NaN,
-      "step": 134,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.2048,
-      "grad_norm": 2.234375,
-      "learning_rate": 2.4597777256685556e-05,
-      "long_answer_loss": 0.2048,
-      "loss": 0.1948,
-      "short_answer_loss": NaN,
-      "step": 135,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1825,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.45894871821082e-05,
-      "long_answer_loss": 0.1825,
-      "loss": 0.1836,
-      "short_answer_loss": NaN,
-      "step": 136,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1889,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.45811139706475e-05,
-      "long_answer_loss": 0.1889,
-      "loss": 0.1848,
-      "short_answer_loss": NaN,
-      "step": 137,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.1961,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.4572657679884285e-05,
-      "long_answer_loss": 0.1961,
-      "loss": 0.179,
-      "short_answer_loss": NaN,
-      "step": 138,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.22,
-      "full_loss": 0.197,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.4564118367970706e-05,
-      "long_answer_loss": 0.197,
-      "loss": 0.1927,
-      "short_answer_loss": NaN,
-      "step": 139,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.2099,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.455549609362983e-05,
-      "long_answer_loss": 0.2099,
-      "loss": 0.1934,
-      "short_answer_loss": NaN,
-      "step": 140,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1903,
-      "grad_norm": 2.28125,
-      "learning_rate": 2.4546790916155243e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1921,
-      "short_answer_loss": NaN,
-      "step": 141,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1882,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.4538002895410634e-05,
-      "long_answer_loss": 0.1882,
-      "loss": 0.1834,
-      "short_answer_loss": NaN,
-      "step": 142,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.2116,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.452913209182939e-05,
-      "long_answer_loss": 0.2116,
-      "loss": 0.1767,
-      "short_answer_loss": NaN,
-      "step": 143,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1931,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.4520178566414177e-05,
-      "long_answer_loss": 0.1931,
-      "loss": 0.1955,
-      "short_answer_loss": NaN,
-      "step": 144,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.23,
-      "full_loss": 0.1908,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.4511142380736517e-05,
-      "long_answer_loss": 0.1908,
-      "loss": 0.1934,
-      "short_answer_loss": NaN,
-      "step": 145,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2259,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.450202359693639e-05,
-      "long_answer_loss": 0.2259,
-      "loss": 0.1822,
-      "short_answer_loss": NaN,
-      "step": 146,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.2343,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.449282227772176e-05,
-      "long_answer_loss": 0.2343,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 147,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.1845,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.4483538486368186e-05,
-      "long_answer_loss": 0.1845,
-      "loss": 0.1914,
-      "short_answer_loss": NaN,
-      "step": 148,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.1773,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.4474172286718363e-05,
-      "long_answer_loss": 0.1773,
-      "loss": 0.179,
-      "short_answer_loss": NaN,
-      "step": 149,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.187,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.4464723743181693e-05,
-      "long_answer_loss": 0.187,
-      "loss": 0.1869,
-      "short_answer_loss": NaN,
-      "step": 150,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.24,
-      "full_loss": 0.1795,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.445519292073385e-05,
-      "long_answer_loss": 0.1795,
-      "loss": 0.1928,
-      "short_answer_loss": NaN,
-      "step": 151,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1942,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.4445579884916297e-05,
-      "long_answer_loss": 0.1942,
-      "loss": 0.1897,
-      "short_answer_loss": NaN,
-      "step": 152,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.183,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.443588470183589e-05,
-      "long_answer_loss": 0.183,
-      "loss": 0.1922,
-      "short_answer_loss": NaN,
-      "step": 153,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1974,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.442610743816438e-05,
-      "long_answer_loss": 0.1974,
-      "loss": 0.1998,
-      "short_answer_loss": NaN,
-      "step": 154,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.169,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.4416248161137972e-05,
-      "long_answer_loss": 0.169,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 155,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1794,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.4406306938556853e-05,
-      "long_answer_loss": 0.1794,
-      "loss": 0.1882,
-      "short_answer_loss": NaN,
-      "step": 156,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.25,
-      "full_loss": 0.1685,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.4396283838784743e-05,
-      "long_answer_loss": 0.1685,
-      "loss": 0.1772,
-      "short_answer_loss": NaN,
-      "step": 157,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1855,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.438617893074841e-05,
-      "long_answer_loss": 0.1855,
-      "loss": 0.1951,
-      "short_answer_loss": NaN,
-      "step": 158,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1648,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.4375992283937194e-05,
-      "long_answer_loss": 0.1648,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 159,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1721,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.4365723968402552e-05,
-      "long_answer_loss": 0.1721,
-      "loss": 0.1834,
-      "short_answer_loss": NaN,
-      "step": 160,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1973,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.4355374054757546e-05,
-      "long_answer_loss": 0.1973,
-      "loss": 0.1833,
-      "short_answer_loss": NaN,
-      "step": 161,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1963,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.434494261417637e-05,
-      "long_answer_loss": 0.1963,
-      "loss": 0.1867,
-      "short_answer_loss": NaN,
-      "step": 162,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.26,
-      "full_loss": 0.1969,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.433442971839387e-05,
-      "long_answer_loss": 0.1969,
-      "loss": 0.1884,
-      "short_answer_loss": NaN,
-      "step": 163,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.2025,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.432383543970504e-05,
-      "long_answer_loss": 0.2025,
-      "loss": 0.1926,
-      "short_answer_loss": NaN,
-      "step": 164,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.2021,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.4313159850964523e-05,
-      "long_answer_loss": 0.2021,
-      "loss": 0.175,
-      "short_answer_loss": NaN,
-      "step": 165,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.159,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.4302403025586122e-05,
-      "long_answer_loss": 0.159,
-      "loss": 0.1889,
-      "short_answer_loss": NaN,
-      "step": 166,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1843,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.429156503754228e-05,
-      "long_answer_loss": 0.1843,
-      "loss": 0.1956,
-      "short_answer_loss": NaN,
-      "step": 167,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1944,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.428064596136358e-05,
-      "long_answer_loss": 0.1944,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 168,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1753,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.4269645872138237e-05,
-      "long_answer_loss": 0.1753,
-      "loss": 0.1751,
-      "short_answer_loss": NaN,
-      "step": 169,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.27,
-      "full_loss": 0.1986,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.4258564845511568e-05,
-      "long_answer_loss": 0.1986,
-      "loss": 0.1929,
-      "short_answer_loss": NaN,
-      "step": 170,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1973,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.4247402957685482e-05,
-      "long_answer_loss": 0.1973,
-      "loss": 0.1852,
-      "short_answer_loss": NaN,
-      "step": 171,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.2181,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4236160285417964e-05,
-      "long_answer_loss": 0.2181,
-      "loss": 0.2004,
-      "short_answer_loss": NaN,
-      "step": 172,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1811,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.4224836906022518e-05,
-      "long_answer_loss": 0.1811,
-      "loss": 0.1879,
-      "short_answer_loss": NaN,
-      "step": 173,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1941,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.421343289736767e-05,
-      "long_answer_loss": 0.1941,
-      "loss": 0.1896,
-      "short_answer_loss": NaN,
-      "step": 174,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.1872,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.4201948337876405e-05,
-      "long_answer_loss": 0.1872,
-      "loss": 0.1869,
-      "short_answer_loss": NaN,
-      "step": 175,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.28,
-      "full_loss": 0.2018,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4190383306525647e-05,
-      "long_answer_loss": 0.2018,
-      "loss": 0.188,
-      "short_answer_loss": NaN,
-      "step": 176,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1861,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.4178737882845708e-05,
-      "long_answer_loss": 0.1861,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 177,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2307,
-      "grad_norm": 2.0,
-      "learning_rate": 2.4167012146919735e-05,
-      "long_answer_loss": 0.2307,
-      "loss": 0.1893,
-      "short_answer_loss": NaN,
-      "step": 178,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1916,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.4155206179383172e-05,
-      "long_answer_loss": 0.1916,
-      "loss": 0.1842,
-      "short_answer_loss": NaN,
-      "step": 179,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.2011,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.41433200614232e-05,
-      "long_answer_loss": 0.2011,
-      "loss": 0.192,
-      "short_answer_loss": NaN,
-      "step": 180,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1894,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4131353874778168e-05,
-      "long_answer_loss": 0.1894,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 181,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.29,
-      "full_loss": 0.1915,
-      "grad_norm": 2.0625,
-      "learning_rate": 2.4119307701737053e-05,
-      "long_answer_loss": 0.1915,
-      "loss": 0.196,
-      "short_answer_loss": NaN,
-      "step": 182,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2175,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.4107181625138874e-05,
-      "long_answer_loss": 0.2175,
-      "loss": 0.1841,
-      "short_answer_loss": NaN,
-      "step": 183,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.1877,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.4094975728372133e-05,
-      "long_answer_loss": 0.1877,
-      "loss": 0.1905,
-      "short_answer_loss": NaN,
-      "step": 184,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2038,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.4082690095374234e-05,
-      "long_answer_loss": 0.2038,
-      "loss": 0.1906,
-      "short_answer_loss": NaN,
-      "step": 185,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2006,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.407032481063092e-05,
-      "long_answer_loss": 0.2006,
-      "loss": 0.1848,
-      "short_answer_loss": NaN,
-      "step": 186,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.2446,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.4057879959175672e-05,
-      "long_answer_loss": 0.2446,
-      "loss": 0.196,
-      "short_answer_loss": NaN,
-      "step": 187,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.3,
-      "full_loss": 0.199,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.4045355626589145e-05,
-      "long_answer_loss": 0.199,
-      "loss": 0.1846,
-      "short_answer_loss": NaN,
-      "step": 188,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.2025,
-      "grad_norm": 2.265625,
-      "learning_rate": 2.4032751898998555e-05,
-      "long_answer_loss": 0.2025,
-      "loss": 0.1986,
-      "short_answer_loss": NaN,
-      "step": 189,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1821,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.4020068863077116e-05,
-      "long_answer_loss": 0.1821,
-      "loss": 0.1906,
-      "short_answer_loss": NaN,
-      "step": 190,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1585,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.4007306606043416e-05,
-      "long_answer_loss": 0.1585,
-      "loss": 0.1821,
-      "short_answer_loss": NaN,
-      "step": 191,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1971,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.3994465215660846e-05,
-      "long_answer_loss": 0.1971,
-      "loss": 0.1865,
-      "short_answer_loss": NaN,
-      "step": 192,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1936,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.3981544780236963e-05,
-      "long_answer_loss": 0.1936,
-      "loss": 0.1786,
-      "short_answer_loss": NaN,
-      "step": 193,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.31,
-      "full_loss": 0.1601,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.3968545388622917e-05,
-      "long_answer_loss": 0.1601,
-      "loss": 0.184,
-      "short_answer_loss": NaN,
-      "step": 194,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1794,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.39554671302128e-05,
-      "long_answer_loss": 0.1794,
-      "loss": 0.1826,
-      "short_answer_loss": NaN,
-      "step": 195,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2061,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.3942310094943083e-05,
-      "long_answer_loss": 0.2061,
-      "loss": 0.1856,
-      "short_answer_loss": NaN,
-      "step": 196,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1974,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.3929074373291946e-05,
-      "long_answer_loss": 0.1974,
-      "loss": 0.188,
-      "short_answer_loss": NaN,
-      "step": 197,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1961,
-      "grad_norm": 2.109375,
-      "learning_rate": 2.391576005627869e-05,
-      "long_answer_loss": 0.1961,
-      "loss": 0.1941,
-      "short_answer_loss": NaN,
-      "step": 198,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.1659,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.3902367235463104e-05,
-      "long_answer_loss": 0.1659,
-      "loss": 0.1898,
-      "short_answer_loss": NaN,
-      "step": 199,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.221,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.3888896002944815e-05,
-      "long_answer_loss": 0.221,
-      "loss": 0.195,
-      "short_answer_loss": NaN,
-      "step": 200,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.32,
-      "full_loss": 0.2119,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.387534645136269e-05,
-      "long_answer_loss": 0.2119,
-      "loss": 0.1854,
-      "short_answer_loss": NaN,
-      "step": 201,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.2143,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.3861718673894166e-05,
-      "long_answer_loss": 0.2143,
-      "loss": 0.1867,
-      "short_answer_loss": NaN,
-      "step": 202,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1887,
-      "grad_norm": 2.125,
-      "learning_rate": 2.384801276425463e-05,
-      "long_answer_loss": 0.1887,
-      "loss": 0.1929,
-      "short_answer_loss": NaN,
-      "step": 203,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1986,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.3834228816696763e-05,
-      "long_answer_loss": 0.1986,
-      "loss": 0.1885,
-      "short_answer_loss": NaN,
-      "step": 204,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.2079,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.3820366926009903e-05,
-      "long_answer_loss": 0.2079,
-      "loss": 0.1958,
-      "short_answer_loss": NaN,
-      "step": 205,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1936,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.3806427187519376e-05,
-      "long_answer_loss": 0.1936,
-      "loss": 0.1938,
-      "short_answer_loss": NaN,
-      "step": 206,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.33,
-      "full_loss": 0.1903,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.3792409697085866e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1859,
-      "short_answer_loss": NaN,
-      "step": 207,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1969,
-      "grad_norm": 2.09375,
-      "learning_rate": 2.3778314551104725e-05,
-      "long_answer_loss": 0.1969,
-      "loss": 0.1914,
-      "short_answer_loss": NaN,
-      "step": 208,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1905,
-      "grad_norm": 1.9921875,
-      "learning_rate": 2.376414184650534e-05,
-      "long_answer_loss": 0.1905,
-      "loss": 0.1832,
-      "short_answer_loss": NaN,
-      "step": 209,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1684,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.3749891680750445e-05,
-      "long_answer_loss": 0.1684,
-      "loss": 0.1765,
-      "short_answer_loss": NaN,
-      "step": 210,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1781,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.3735564151835462e-05,
-      "long_answer_loss": 0.1781,
-      "loss": 0.1868,
-      "short_answer_loss": NaN,
-      "step": 211,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1583,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.3721159358287815e-05,
-      "long_answer_loss": 0.1583,
-      "loss": 0.1739,
-      "short_answer_loss": NaN,
-      "step": 212,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.34,
-      "full_loss": 0.1895,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.370667739916627e-05,
-      "long_answer_loss": 0.1895,
-      "loss": 0.1963,
-      "short_answer_loss": NaN,
-      "step": 213,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1893,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.369211837406024e-05,
-      "long_answer_loss": 0.1893,
-      "loss": 0.1824,
-      "short_answer_loss": NaN,
-      "step": 214,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.177,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.3677482383089105e-05,
-      "long_answer_loss": 0.177,
-      "loss": 0.1841,
-      "short_answer_loss": NaN,
-      "step": 215,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1768,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.3662769526901526e-05,
-      "long_answer_loss": 0.1768,
-      "loss": 0.1736,
-      "short_answer_loss": NaN,
-      "step": 216,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1741,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.364797990667475e-05,
-      "long_answer_loss": 0.1741,
-      "loss": 0.1913,
-      "short_answer_loss": NaN,
-      "step": 217,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1897,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.3633113624113908e-05,
-      "long_answer_loss": 0.1897,
-      "loss": 0.1854,
-      "short_answer_loss": NaN,
-      "step": 218,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.35,
-      "full_loss": 0.1796,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.3618170781451328e-05,
-      "long_answer_loss": 0.1796,
-      "loss": 0.1837,
-      "short_answer_loss": NaN,
-      "step": 219,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1461,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.3603151481445823e-05,
-      "long_answer_loss": 0.1461,
-      "loss": 0.1872,
-      "short_answer_loss": NaN,
-      "step": 220,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1872,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.3588055827381995e-05,
-      "long_answer_loss": 0.1872,
-      "loss": 0.1828,
-      "short_answer_loss": NaN,
-      "step": 221,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1991,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.35728839230695e-05,
-      "long_answer_loss": 0.1991,
-      "loss": 0.1823,
-      "short_answer_loss": NaN,
-      "step": 222,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.2084,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.3557635872842372e-05,
-      "long_answer_loss": 0.2084,
-      "loss": 0.1903,
-      "short_answer_loss": NaN,
-      "step": 223,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.1454,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.3542311781558263e-05,
-      "long_answer_loss": 0.1454,
-      "loss": 0.1738,
-      "short_answer_loss": NaN,
-      "step": 224,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.36,
-      "full_loss": 0.2094,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.3526911754597763e-05,
-      "long_answer_loss": 0.2094,
-      "loss": 0.1725,
-      "short_answer_loss": NaN,
-      "step": 225,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1653,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.3511435897863647e-05,
-      "long_answer_loss": 0.1653,
-      "loss": 0.1774,
-      "short_answer_loss": NaN,
-      "step": 226,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1597,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.3495884317780154e-05,
-      "long_answer_loss": 0.1597,
-      "loss": 0.1845,
-      "short_answer_loss": NaN,
-      "step": 227,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1791,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.3480257121292254e-05,
-      "long_answer_loss": 0.1791,
-      "loss": 0.1814,
-      "short_answer_loss": NaN,
-      "step": 228,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.161,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.3464554415864927e-05,
-      "long_answer_loss": 0.161,
-      "loss": 0.1722,
-      "short_answer_loss": NaN,
-      "step": 229,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.1547,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.3448776309482402e-05,
-      "long_answer_loss": 0.1547,
-      "loss": 0.1762,
-      "short_answer_loss": NaN,
-      "step": 230,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.37,
-      "full_loss": 0.163,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.3432922910647426e-05,
-      "long_answer_loss": 0.163,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 231,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1822,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.341699432838052e-05,
-      "long_answer_loss": 0.1822,
-      "loss": 0.1915,
-      "short_answer_loss": NaN,
-      "step": 232,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1643,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.3400990672219226e-05,
-      "long_answer_loss": 0.1643,
-      "loss": 0.1743,
-      "short_answer_loss": NaN,
-      "step": 233,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1864,
-      "grad_norm": 1.875,
-      "learning_rate": 2.3384912052217345e-05,
-      "long_answer_loss": 0.1864,
-      "loss": 0.1837,
-      "short_answer_loss": NaN,
-      "step": 234,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.2088,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.3368758578944205e-05,
-      "long_answer_loss": 0.2088,
-      "loss": 0.1754,
-      "short_answer_loss": NaN,
-      "step": 235,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1659,
-      "grad_norm": 1.6953125,
-      "learning_rate": 2.3352530363483866e-05,
-      "long_answer_loss": 0.1659,
-      "loss": 0.1826,
-      "short_answer_loss": NaN,
-      "step": 236,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1785,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.3336227517434385e-05,
-      "long_answer_loss": 0.1785,
-      "loss": 0.1885,
-      "short_answer_loss": NaN,
-      "step": 237,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.38,
-      "full_loss": 0.1631,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.331985015290704e-05,
-      "long_answer_loss": 0.1631,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 238,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1705,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.330339838252555e-05,
-      "long_answer_loss": 0.1705,
-      "loss": 0.1753,
-      "short_answer_loss": NaN,
-      "step": 239,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1575,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.3286872319425312e-05,
-      "long_answer_loss": 0.1575,
-      "loss": 0.1864,
-      "short_answer_loss": NaN,
-      "step": 240,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1795,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.3270272077252613e-05,
-      "long_answer_loss": 0.1795,
-      "loss": 0.18,
-      "short_answer_loss": NaN,
-      "step": 241,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1681,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.3253597770163866e-05,
-      "long_answer_loss": 0.1681,
-      "loss": 0.1814,
-      "short_answer_loss": NaN,
-      "step": 242,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1861,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.3236849512824793e-05,
-      "long_answer_loss": 0.1861,
-      "loss": 0.184,
-      "short_answer_loss": NaN,
-      "step": 243,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.39,
-      "full_loss": 0.1786,
-      "grad_norm": 2.0,
-      "learning_rate": 2.322002742040968e-05,
-      "long_answer_loss": 0.1786,
-      "loss": 0.1818,
-      "short_answer_loss": NaN,
-      "step": 244,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1927,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.3203131608600548e-05,
-      "long_answer_loss": 0.1927,
-      "loss": 0.1988,
-      "short_answer_loss": NaN,
-      "step": 245,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1746,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.318616219358637e-05,
-      "long_answer_loss": 0.1746,
-      "loss": 0.1711,
-      "short_answer_loss": NaN,
-      "step": 246,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.159,
-      "grad_norm": 1.875,
-      "learning_rate": 2.3169119292062273e-05,
-      "long_answer_loss": 0.159,
-      "loss": 0.181,
-      "short_answer_loss": NaN,
-      "step": 247,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1809,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.3152003021228746e-05,
-      "long_answer_loss": 0.1809,
-      "loss": 0.1942,
-      "short_answer_loss": NaN,
-      "step": 248,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1779,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.3134813498790814e-05,
-      "long_answer_loss": 0.1779,
-      "loss": 0.1757,
-      "short_answer_loss": NaN,
-      "step": 249,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.4,
-      "full_loss": 0.1589,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.311755084295723e-05,
-      "long_answer_loss": 0.1589,
-      "loss": 0.1693,
-      "short_answer_loss": NaN,
-      "step": 250,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1904,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.3100215172439693e-05,
-      "long_answer_loss": 0.1904,
-      "loss": 0.1824,
-      "short_answer_loss": NaN,
-      "step": 251,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1714,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.308280660645199e-05,
-      "long_answer_loss": 0.1714,
-      "loss": 0.1849,
-      "short_answer_loss": NaN,
-      "step": 252,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.203,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.3065325264709196e-05,
-      "long_answer_loss": 0.203,
-      "loss": 0.1783,
-      "short_answer_loss": NaN,
-      "step": 253,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1694,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.3047771267426866e-05,
-      "long_answer_loss": 0.1694,
-      "loss": 0.1777,
-      "short_answer_loss": NaN,
-      "step": 254,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.2025,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.303014473532017e-05,
-      "long_answer_loss": 0.2025,
-      "loss": 0.1794,
-      "short_answer_loss": NaN,
-      "step": 255,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.41,
-      "full_loss": 0.1735,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.3012445789603093e-05,
-      "long_answer_loss": 0.1735,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 256,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1522,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.2994674551987605e-05,
-      "long_answer_loss": 0.1522,
-      "loss": 0.1789,
-      "short_answer_loss": NaN,
-      "step": 257,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1737,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.2976831144682797e-05,
-      "long_answer_loss": 0.1737,
-      "loss": 0.18,
-      "short_answer_loss": NaN,
-      "step": 258,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1774,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.295891569039406e-05,
-      "long_answer_loss": 0.1774,
-      "loss": 0.1879,
-      "short_answer_loss": NaN,
-      "step": 259,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1798,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.2940928312322246e-05,
-      "long_answer_loss": 0.1798,
-      "loss": 0.1803,
-      "short_answer_loss": NaN,
-      "step": 260,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.2178,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.29228691341628e-05,
-      "long_answer_loss": 0.2178,
-      "loss": 0.1801,
-      "short_answer_loss": NaN,
-      "step": 261,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.42,
-      "full_loss": 0.1985,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.2904738280104927e-05,
-      "long_answer_loss": 0.1985,
-      "loss": 0.1893,
-      "short_answer_loss": NaN,
-      "step": 262,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1651,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.2886535874830726e-05,
-      "long_answer_loss": 0.1651,
-      "loss": 0.1812,
-      "short_answer_loss": NaN,
-      "step": 263,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.2057,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.286826204351435e-05,
-      "long_answer_loss": 0.2057,
-      "loss": 0.1824,
-      "short_answer_loss": NaN,
-      "step": 264,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1897,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.284991691182113e-05,
-      "long_answer_loss": 0.1897,
-      "loss": 0.1896,
-      "short_answer_loss": NaN,
-      "step": 265,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1933,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.2831500605906702e-05,
-      "long_answer_loss": 0.1933,
-      "loss": 0.1734,
-      "short_answer_loss": NaN,
-      "step": 266,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1577,
-      "grad_norm": 1.7109375,
-      "learning_rate": 2.281301325241617e-05,
-      "long_answer_loss": 0.1577,
-      "loss": 0.1691,
-      "short_answer_loss": NaN,
-      "step": 267,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.161,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.279445497848321e-05,
-      "long_answer_loss": 0.161,
-      "loss": 0.181,
-      "short_answer_loss": NaN,
-      "step": 268,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.43,
-      "full_loss": 0.1905,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.2775825911729207e-05,
-      "long_answer_loss": 0.1905,
-      "loss": 0.1961,
-      "short_answer_loss": NaN,
-      "step": 269,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1823,
-      "grad_norm": 1.75,
-      "learning_rate": 2.275712618026236e-05,
-      "long_answer_loss": 0.1823,
-      "loss": 0.1751,
-      "short_answer_loss": NaN,
-      "step": 270,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1653,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.2738355912676838e-05,
-      "long_answer_loss": 0.1653,
-      "loss": 0.1805,
-      "short_answer_loss": NaN,
-      "step": 271,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1818,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.2719515238051846e-05,
-      "long_answer_loss": 0.1818,
-      "loss": 0.1761,
-      "short_answer_loss": NaN,
-      "step": 272,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1863,
-      "grad_norm": 1.984375,
-      "learning_rate": 2.2700604285950783e-05,
-      "long_answer_loss": 0.1863,
-      "loss": 0.1864,
-      "short_answer_loss": NaN,
-      "step": 273,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1573,
-      "grad_norm": 2.078125,
-      "learning_rate": 2.2681623186420323e-05,
-      "long_answer_loss": 0.1573,
-      "loss": 0.1811,
-      "short_answer_loss": NaN,
-      "step": 274,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.44,
-      "full_loss": 0.1858,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.266257206998953e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1731,
-      "short_answer_loss": NaN,
-      "step": 275,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.185,
-      "grad_norm": 2.125,
-      "learning_rate": 2.264345106766896e-05,
-      "long_answer_loss": 0.185,
-      "loss": 0.1823,
-      "short_answer_loss": NaN,
-      "step": 276,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.204,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.2624260310949763e-05,
-      "long_answer_loss": 0.204,
-      "loss": 0.1764,
-      "short_answer_loss": NaN,
-      "step": 277,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.1896,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.2604999931802773e-05,
-      "long_answer_loss": 0.1896,
-      "loss": 0.1822,
-      "short_answer_loss": NaN,
-      "step": 278,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.191,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.25856700626776e-05,
-      "long_answer_loss": 0.191,
-      "loss": 0.1832,
-      "short_answer_loss": NaN,
-      "step": 279,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.1685,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.2566270836501725e-05,
-      "long_answer_loss": 0.1685,
-      "loss": 0.1778,
-      "short_answer_loss": NaN,
-      "step": 280,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.45,
-      "full_loss": 0.171,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.2546802386679585e-05,
-      "long_answer_loss": 0.171,
-      "loss": 0.1776,
-      "short_answer_loss": NaN,
-      "step": 281,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.2089,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.2527264847091652e-05,
-      "long_answer_loss": 0.2089,
-      "loss": 0.1911,
-      "short_answer_loss": NaN,
-      "step": 282,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1788,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.2507658352093503e-05,
-      "long_answer_loss": 0.1788,
-      "loss": 0.1738,
-      "short_answer_loss": NaN,
-      "step": 283,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1825,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.2487983036514932e-05,
-      "long_answer_loss": 0.1825,
-      "loss": 0.179,
-      "short_answer_loss": NaN,
-      "step": 284,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1738,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.2468239035658972e-05,
-      "long_answer_loss": 0.1738,
-      "loss": 0.1746,
-      "short_answer_loss": NaN,
-      "step": 285,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1766,
-      "grad_norm": 1.75,
-      "learning_rate": 2.2448426485301006e-05,
-      "long_answer_loss": 0.1766,
-      "loss": 0.1774,
-      "short_answer_loss": NaN,
-      "step": 286,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.46,
-      "full_loss": 0.1972,
-      "grad_norm": 1.6171875,
-      "learning_rate": 2.2428545521687816e-05,
-      "long_answer_loss": 0.1972,
-      "loss": 0.1653,
-      "short_answer_loss": NaN,
-      "step": 287,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1686,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.2408596281536638e-05,
-      "long_answer_loss": 0.1686,
-      "loss": 0.1737,
-      "short_answer_loss": NaN,
-      "step": 288,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1981,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.2388578902034243e-05,
-      "long_answer_loss": 0.1981,
-      "loss": 0.1735,
-      "short_answer_loss": NaN,
-      "step": 289,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1923,
-      "grad_norm": 1.9453125,
-      "learning_rate": 2.2368493520835977e-05,
-      "long_answer_loss": 0.1923,
-      "loss": 0.1842,
-      "short_answer_loss": NaN,
-      "step": 290,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.2005,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.2348340276064816e-05,
-      "long_answer_loss": 0.2005,
-      "loss": 0.1797,
-      "short_answer_loss": NaN,
-      "step": 291,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.179,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.2328119306310423e-05,
-      "long_answer_loss": 0.179,
-      "loss": 0.1784,
-      "short_answer_loss": NaN,
-      "step": 292,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.47,
-      "full_loss": 0.1715,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.23078307506282e-05,
-      "long_answer_loss": 0.1715,
-      "loss": 0.1694,
-      "short_answer_loss": NaN,
-      "step": 293,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1441,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.2287474748538308e-05,
-      "long_answer_loss": 0.1441,
-      "loss": 0.1689,
-      "short_answer_loss": NaN,
-      "step": 294,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1673,
-      "grad_norm": 1.875,
-      "learning_rate": 2.2267051440024734e-05,
-      "long_answer_loss": 0.1673,
-      "loss": 0.178,
-      "short_answer_loss": NaN,
-      "step": 295,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.2067,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.2246560965534312e-05,
-      "long_answer_loss": 0.2067,
-      "loss": 0.1841,
-      "short_answer_loss": NaN,
-      "step": 296,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1729,
-      "grad_norm": 1.75,
-      "learning_rate": 2.222600346597576e-05,
-      "long_answer_loss": 0.1729,
-      "loss": 0.1665,
-      "short_answer_loss": NaN,
-      "step": 297,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1546,
-      "grad_norm": 1.921875,
-      "learning_rate": 2.2205379082718725e-05,
-      "long_answer_loss": 0.1546,
-      "loss": 0.1866,
-      "short_answer_loss": NaN,
-      "step": 298,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.48,
-      "full_loss": 0.1767,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.2184687957592786e-05,
-      "long_answer_loss": 0.1767,
-      "loss": 0.1773,
-      "short_answer_loss": NaN,
-      "step": 299,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1712,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.21639302328865e-05,
-      "long_answer_loss": 0.1712,
-      "loss": 0.1779,
-      "short_answer_loss": NaN,
-      "step": 300,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1846,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.2143106051346407e-05,
-      "long_answer_loss": 0.1846,
-      "loss": 0.1862,
-      "short_answer_loss": NaN,
-      "step": 301,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1772,
-      "grad_norm": 1.8671875,
-      "learning_rate": 2.2122215556176074e-05,
-      "long_answer_loss": 0.1772,
-      "loss": 0.1744,
-      "short_answer_loss": NaN,
-      "step": 302,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.172,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.2101258891035075e-05,
-      "long_answer_loss": 0.172,
-      "loss": 0.1764,
-      "short_answer_loss": NaN,
-      "step": 303,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1873,
-      "grad_norm": 1.9609375,
-      "learning_rate": 2.2080236200038026e-05,
-      "long_answer_loss": 0.1873,
-      "loss": 0.1825,
-      "short_answer_loss": NaN,
-      "step": 304,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1944,
-      "grad_norm": 1.96875,
-      "learning_rate": 2.2059147627753595e-05,
-      "long_answer_loss": 0.1944,
-      "loss": 0.1785,
-      "short_answer_loss": NaN,
-      "step": 305,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.49,
-      "full_loss": 0.1921,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.2037993319203498e-05,
-      "long_answer_loss": 0.1921,
-      "loss": 0.1716,
-      "short_answer_loss": NaN,
-      "step": 306,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1942,
-      "grad_norm": 1.9765625,
-      "learning_rate": 2.20167734198615e-05,
-      "long_answer_loss": 0.1942,
-      "loss": 0.1826,
-      "short_answer_loss": NaN,
-      "step": 307,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1896,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.1995488075652433e-05,
-      "long_answer_loss": 0.1896,
-      "loss": 0.1821,
-      "short_answer_loss": NaN,
-      "step": 308,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1871,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.1974137432951165e-05,
-      "long_answer_loss": 0.1871,
-      "loss": 0.1877,
-      "short_answer_loss": NaN,
-      "step": 309,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1826,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.195272163858162e-05,
-      "long_answer_loss": 0.1826,
-      "loss": 0.1769,
-      "short_answer_loss": NaN,
-      "step": 310,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1859,
-      "grad_norm": 2.03125,
-      "learning_rate": 2.193124083981575e-05,
-      "long_answer_loss": 0.1859,
-      "loss": 0.1716,
-      "short_answer_loss": NaN,
-      "step": 311,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.5,
-      "full_loss": 0.1858,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.190969518437253e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1744,
-      "short_answer_loss": NaN,
-      "step": 312,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1558,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.1888084820416944e-05,
-      "long_answer_loss": 0.1558,
-      "loss": 0.1725,
-      "short_answer_loss": NaN,
-      "step": 313,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1731,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.186640989655896e-05,
-      "long_answer_loss": 0.1731,
-      "loss": 0.189,
-      "short_answer_loss": NaN,
-      "step": 314,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1705,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.1844670561852508e-05,
-      "long_answer_loss": 0.1705,
-      "loss": 0.1745,
-      "short_answer_loss": NaN,
-      "step": 315,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1566,
-      "grad_norm": 2.046875,
-      "learning_rate": 2.1822866965794465e-05,
-      "long_answer_loss": 0.1566,
-      "loss": 0.1757,
-      "short_answer_loss": NaN,
-      "step": 316,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1705,
-      "grad_norm": 1.953125,
-      "learning_rate": 2.180099925832361e-05,
-      "long_answer_loss": 0.1705,
-      "loss": 0.173,
-      "short_answer_loss": NaN,
-      "step": 317,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.51,
-      "full_loss": 0.1721,
-      "grad_norm": 1.8828125,
-      "learning_rate": 2.177906758981962e-05,
-      "long_answer_loss": 0.1721,
-      "loss": 0.1692,
-      "short_answer_loss": NaN,
-      "step": 318,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1687,
-      "grad_norm": 2.015625,
-      "learning_rate": 2.1757072111101994e-05,
-      "long_answer_loss": 0.1687,
-      "loss": 0.1773,
-      "short_answer_loss": NaN,
-      "step": 319,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1731,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.1735012973429068e-05,
-      "long_answer_loss": 0.1731,
-      "loss": 0.1768,
-      "short_answer_loss": NaN,
-      "step": 320,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1698,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.1712890328496927e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.1804,
-      "short_answer_loss": NaN,
-      "step": 321,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1582,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.1690704328438384e-05,
-      "long_answer_loss": 0.1582,
-      "loss": 0.1813,
-      "short_answer_loss": NaN,
-      "step": 322,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.1814,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1668455125821945e-05,
-      "long_answer_loss": 0.1814,
-      "loss": 0.1804,
-      "short_answer_loss": NaN,
-      "step": 323,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.52,
-      "full_loss": 0.2042,
-      "grad_norm": 1.9296875,
-      "learning_rate": 2.1646142873650738e-05,
-      "long_answer_loss": 0.2042,
-      "loss": 0.1828,
-      "short_answer_loss": NaN,
-      "step": 324,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1851,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.1623767725361466e-05,
-      "long_answer_loss": 0.1851,
-      "loss": 0.1888,
-      "short_answer_loss": NaN,
-      "step": 325,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1432,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.160132983482336e-05,
-      "long_answer_loss": 0.1432,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 326,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1712,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.157882935633712e-05,
-      "long_answer_loss": 0.1712,
-      "loss": 0.1735,
-      "short_answer_loss": NaN,
-      "step": 327,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1751,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.1556266444633845e-05,
-      "long_answer_loss": 0.1751,
-      "loss": 0.1697,
-      "short_answer_loss": NaN,
-      "step": 328,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1779,
-      "grad_norm": 1.7578125,
-      "learning_rate": 2.153364125487397e-05,
-      "long_answer_loss": 0.1779,
-      "loss": 0.1869,
-      "short_answer_loss": NaN,
-      "step": 329,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.53,
-      "full_loss": 0.1904,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.1510953942646215e-05,
-      "long_answer_loss": 0.1904,
-      "loss": 0.1821,
-      "short_answer_loss": NaN,
-      "step": 330,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1818,
-      "grad_norm": 1.859375,
-      "learning_rate": 2.1488204663966498e-05,
-      "long_answer_loss": 0.1818,
-      "loss": 0.1789,
-      "short_answer_loss": NaN,
-      "step": 331,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1572,
-      "grad_norm": 1.6953125,
-      "learning_rate": 2.1465393575276867e-05,
-      "long_answer_loss": 0.1572,
-      "loss": 0.1721,
-      "short_answer_loss": NaN,
-      "step": 332,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1797,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.1442520833444416e-05,
-      "long_answer_loss": 0.1797,
-      "loss": 0.1697,
-      "short_answer_loss": NaN,
-      "step": 333,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1511,
-      "grad_norm": 1.7421875,
-      "learning_rate": 2.1419586595760226e-05,
-      "long_answer_loss": 0.1511,
-      "loss": 0.1778,
-      "short_answer_loss": NaN,
-      "step": 334,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1576,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.1396591019938278e-05,
-      "long_answer_loss": 0.1576,
-      "loss": 0.1691,
-      "short_answer_loss": NaN,
-      "step": 335,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1842,
-      "grad_norm": 1.5859375,
-      "learning_rate": 2.1373534264114344e-05,
-      "long_answer_loss": 0.1842,
-      "loss": 0.1698,
-      "short_answer_loss": NaN,
-      "step": 336,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.54,
-      "full_loss": 0.1586,
-      "grad_norm": 1.734375,
-      "learning_rate": 2.1350416486844928e-05,
-      "long_answer_loss": 0.1586,
-      "loss": 0.1704,
-      "short_answer_loss": NaN,
-      "step": 337,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1903,
-      "grad_norm": 1.7109375,
-      "learning_rate": 2.1327237847106167e-05,
-      "long_answer_loss": 0.1903,
-      "loss": 0.1772,
-      "short_answer_loss": NaN,
-      "step": 338,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.161,
-      "grad_norm": 1.6328125,
-      "learning_rate": 2.130399850429274e-05,
-      "long_answer_loss": 0.161,
-      "loss": 0.1591,
-      "short_answer_loss": NaN,
-      "step": 339,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1558,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.1280698618216757e-05,
-      "long_answer_loss": 0.1558,
-      "loss": 0.1735,
-      "short_answer_loss": NaN,
-      "step": 340,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1741,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.125733834910668e-05,
-      "long_answer_loss": 0.1741,
-      "loss": 0.169,
-      "short_answer_loss": NaN,
-      "step": 341,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1632,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.1233917857606212e-05,
-      "long_answer_loss": 0.1632,
-      "loss": 0.1762,
-      "short_answer_loss": NaN,
-      "step": 342,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.55,
-      "full_loss": 0.1648,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.1210437304773185e-05,
-      "long_answer_loss": 0.1648,
-      "loss": 0.1732,
-      "short_answer_loss": NaN,
-      "step": 343,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1699,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.1186896852078476e-05,
-      "long_answer_loss": 0.1699,
-      "loss": 0.176,
-      "short_answer_loss": NaN,
-      "step": 344,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1685,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.1163296661404864e-05,
-      "long_answer_loss": 0.1685,
-      "loss": 0.1694,
-      "short_answer_loss": NaN,
-      "step": 345,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1652,
-      "grad_norm": 1.8125,
-      "learning_rate": 2.113963689504594e-05,
-      "long_answer_loss": 0.1652,
-      "loss": 0.167,
-      "short_answer_loss": NaN,
-      "step": 346,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1597,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.111591771570499e-05,
-      "long_answer_loss": 0.1597,
-      "loss": 0.1695,
-      "short_answer_loss": NaN,
-      "step": 347,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.2015,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.1092139286493866e-05,
-      "long_answer_loss": 0.2015,
-      "loss": 0.1803,
-      "short_answer_loss": NaN,
-      "step": 348,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.56,
-      "full_loss": 0.1709,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.106830177093187e-05,
-      "long_answer_loss": 0.1709,
-      "loss": 0.1764,
-      "short_answer_loss": NaN,
-      "step": 349,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.2019,
-      "grad_norm": 1.703125,
-      "learning_rate": 2.104440533294462e-05,
-      "long_answer_loss": 0.2019,
-      "loss": 0.1725,
-      "short_answer_loss": NaN,
-      "step": 350,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1753,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.1020450136862953e-05,
-      "long_answer_loss": 0.1753,
-      "loss": 0.173,
-      "short_answer_loss": NaN,
-      "step": 351,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.18,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.0996436347421744e-05,
-      "long_answer_loss": 0.18,
-      "loss": 0.1768,
-      "short_answer_loss": NaN,
-      "step": 352,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1674,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.0972364129758825e-05,
-      "long_answer_loss": 0.1674,
-      "loss": 0.1755,
-      "short_answer_loss": NaN,
-      "step": 353,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1793,
-      "grad_norm": 1.7734375,
-      "learning_rate": 2.0948233649413815e-05,
-      "long_answer_loss": 0.1793,
-      "loss": 0.1821,
-      "short_answer_loss": NaN,
-      "step": 354,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.57,
-      "full_loss": 0.1591,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.0924045072327003e-05,
-      "long_answer_loss": 0.1591,
-      "loss": 0.1628,
-      "short_answer_loss": NaN,
-      "step": 355,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1624,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.089979856483819e-05,
-      "long_answer_loss": 0.1624,
-      "loss": 0.1692,
-      "short_answer_loss": NaN,
-      "step": 356,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1785,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.0875494293685548e-05,
-      "long_answer_loss": 0.1785,
-      "loss": 0.1717,
-      "short_answer_loss": NaN,
-      "step": 357,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1436,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.0851132426004492e-05,
-      "long_answer_loss": 0.1436,
-      "loss": 0.1697,
-      "short_answer_loss": NaN,
-      "step": 358,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1671,
-      "grad_norm": 1.8203125,
-      "learning_rate": 2.082671312932651e-05,
-      "long_answer_loss": 0.1671,
-      "loss": 0.1679,
-      "short_answer_loss": NaN,
-      "step": 359,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.1741,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.0802236571578e-05,
-      "long_answer_loss": 0.1741,
-      "loss": 0.1639,
-      "short_answer_loss": NaN,
-      "step": 360,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.58,
-      "full_loss": 0.179,
-      "grad_norm": 1.8046875,
-      "learning_rate": 2.0777702921079163e-05,
-      "long_answer_loss": 0.179,
-      "loss": 0.1703,
-      "short_answer_loss": NaN,
-      "step": 361,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1563,
-      "grad_norm": 1.90625,
-      "learning_rate": 2.075311234654279e-05,
-      "long_answer_loss": 0.1563,
-      "loss": 0.1645,
-      "short_answer_loss": NaN,
-      "step": 362,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1723,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.072846501707314e-05,
-      "long_answer_loss": 0.1723,
-      "loss": 0.1693,
-      "short_answer_loss": NaN,
-      "step": 363,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1928,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.0703761102164764e-05,
-      "long_answer_loss": 0.1928,
-      "loss": 0.164,
-      "short_answer_loss": NaN,
-      "step": 364,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1731,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.0679000771701326e-05,
-      "long_answer_loss": 0.1731,
-      "loss": 0.1676,
-      "short_answer_loss": NaN,
-      "step": 365,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1856,
-      "grad_norm": 1.8515625,
-      "learning_rate": 2.0654184195954465e-05,
-      "long_answer_loss": 0.1856,
-      "loss": 0.1661,
-      "short_answer_loss": NaN,
-      "step": 366,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.1614,
-      "grad_norm": 1.6875,
-      "learning_rate": 2.0629311545582598e-05,
-      "long_answer_loss": 0.1614,
-      "loss": 0.1805,
-      "short_answer_loss": NaN,
-      "step": 367,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.59,
-      "full_loss": 0.172,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.0604382991629755e-05,
-      "long_answer_loss": 0.172,
-      "loss": 0.1736,
-      "short_answer_loss": NaN,
-      "step": 368,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1353,
-      "grad_norm": 1.875,
-      "learning_rate": 2.057939870552441e-05,
-      "long_answer_loss": 0.1353,
-      "loss": 0.1733,
-      "short_answer_loss": NaN,
-      "step": 369,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1488,
-      "grad_norm": 1.75,
-      "learning_rate": 2.0554358859078284e-05,
-      "long_answer_loss": 0.1488,
-      "loss": 0.1699,
-      "short_answer_loss": NaN,
-      "step": 370,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1754,
-      "grad_norm": 1.6484375,
-      "learning_rate": 2.0529263624485183e-05,
-      "long_answer_loss": 0.1754,
-      "loss": 0.1711,
-      "short_answer_loss": NaN,
-      "step": 371,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1752,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.0504113174319812e-05,
-      "long_answer_loss": 0.1752,
-      "loss": 0.1788,
-      "short_answer_loss": NaN,
-      "step": 372,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1742,
-      "grad_norm": 1.75,
-      "learning_rate": 2.0478907681536564e-05,
-      "long_answer_loss": 0.1742,
-      "loss": 0.1839,
-      "short_answer_loss": NaN,
-      "step": 373,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.6,
-      "full_loss": 0.1508,
-      "grad_norm": 1.71875,
-      "learning_rate": 2.0453647319468368e-05,
-      "long_answer_loss": 0.1508,
-      "loss": 0.1696,
-      "short_answer_loss": NaN,
-      "step": 374,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1675,
-      "grad_norm": 1.7265625,
-      "learning_rate": 2.0428332261825456e-05,
-      "long_answer_loss": 0.1675,
-      "loss": 0.1737,
-      "short_answer_loss": NaN,
-      "step": 375,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1773,
-      "grad_norm": 1.8984375,
-      "learning_rate": 2.0402962682694214e-05,
-      "long_answer_loss": 0.1773,
-      "loss": 0.1843,
-      "short_answer_loss": NaN,
-      "step": 376,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.19,
-      "grad_norm": 1.9375,
-      "learning_rate": 2.0377538756535947e-05,
-      "long_answer_loss": 0.19,
-      "loss": 0.1772,
-      "short_answer_loss": NaN,
-      "step": 377,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1942,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.0352060658185696e-05,
-      "long_answer_loss": 0.1942,
-      "loss": 0.1824,
-      "short_answer_loss": NaN,
-      "step": 378,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1905,
-      "grad_norm": 1.9140625,
-      "learning_rate": 2.0326528562851028e-05,
-      "long_answer_loss": 0.1905,
-      "loss": 0.1717,
-      "short_answer_loss": NaN,
-      "step": 379,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.61,
-      "full_loss": 0.1579,
-      "grad_norm": 1.84375,
-      "learning_rate": 2.030094264611084e-05,
-      "long_answer_loss": 0.1579,
-      "loss": 0.1621,
-      "short_answer_loss": NaN,
-      "step": 380,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1883,
-      "grad_norm": 1.796875,
-      "learning_rate": 2.027530308391416e-05,
-      "long_answer_loss": 0.1883,
-      "loss": 0.1653,
-      "short_answer_loss": NaN,
-      "step": 381,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1601,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.02496100525789e-05,
-      "long_answer_loss": 0.1601,
-      "loss": 0.1722,
-      "short_answer_loss": NaN,
-      "step": 382,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1761,
-      "grad_norm": 1.828125,
-      "learning_rate": 2.0223863728790682e-05,
-      "long_answer_loss": 0.1761,
-      "loss": 0.1696,
-      "short_answer_loss": NaN,
-      "step": 383,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.192,
-      "grad_norm": 1.78125,
-      "learning_rate": 2.0198064289601615e-05,
-      "long_answer_loss": 0.192,
-      "loss": 0.166,
-      "short_answer_loss": NaN,
-      "step": 384,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.2026,
-      "grad_norm": 1.890625,
-      "learning_rate": 2.017221191242906e-05,
-      "long_answer_loss": 0.2026,
-      "loss": 0.1795,
-      "short_answer_loss": NaN,
-      "step": 385,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.62,
-      "full_loss": 0.1869,
-      "grad_norm": 1.8359375,
-      "learning_rate": 2.014630677505443e-05,
-      "long_answer_loss": 0.1869,
-      "loss": 0.1652,
-      "short_answer_loss": NaN,
-      "step": 386,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1679,
-      "grad_norm": 1.625,
-      "learning_rate": 2.0120349055621952e-05,
-      "long_answer_loss": 0.1679,
-      "loss": 0.173,
-      "short_answer_loss": NaN,
-      "step": 387,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1758,
-      "grad_norm": 1.671875,
-      "learning_rate": 2.0094338932637447e-05,
-      "long_answer_loss": 0.1758,
-      "loss": 0.1693,
-      "short_answer_loss": NaN,
-      "step": 388,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1879,
-      "grad_norm": 1.671875,
-      "learning_rate": 2.0068276584967113e-05,
-      "long_answer_loss": 0.1879,
-      "loss": 0.1621,
-      "short_answer_loss": NaN,
-      "step": 389,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1635,
-      "grad_norm": 1.765625,
-      "learning_rate": 2.0042162191836285e-05,
-      "long_answer_loss": 0.1635,
-      "loss": 0.1748,
-      "short_answer_loss": NaN,
-      "step": 390,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1437,
-      "grad_norm": 1.7890625,
-      "learning_rate": 2.00159959328282e-05,
-      "long_answer_loss": 0.1437,
-      "loss": 0.1693,
-      "short_answer_loss": NaN,
-      "step": 391,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.63,
-      "full_loss": 0.1727,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.9989777987882763e-05,
-      "long_answer_loss": 0.1727,
-      "loss": 0.1669,
-      "short_answer_loss": NaN,
-      "step": 392,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1738,
-      "grad_norm": 1.75,
-      "learning_rate": 1.996350853729532e-05,
-      "long_answer_loss": 0.1738,
-      "loss": 0.1747,
-      "short_answer_loss": NaN,
-      "step": 393,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1698,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.993718776171541e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.1653,
-      "short_answer_loss": NaN,
-      "step": 394,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1624,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.9910815842145513e-05,
-      "long_answer_loss": 0.1624,
-      "loss": 0.1777,
-      "short_answer_loss": NaN,
-      "step": 395,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1493,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.9884392959939824e-05,
-      "long_answer_loss": 0.1493,
-      "loss": 0.1654,
-      "short_answer_loss": NaN,
-      "step": 396,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.166,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.9857919296803e-05,
-      "long_answer_loss": 0.166,
-      "loss": 0.1705,
-      "short_answer_loss": NaN,
-      "step": 397,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.64,
-      "full_loss": 0.1814,
-      "grad_norm": 1.890625,
-      "learning_rate": 1.9831395034788904e-05,
-      "long_answer_loss": 0.1814,
-      "loss": 0.1733,
-      "short_answer_loss": NaN,
-      "step": 398,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.161,
-      "grad_norm": 1.90625,
-      "learning_rate": 1.9804820356299356e-05,
-      "long_answer_loss": 0.161,
-      "loss": 0.1688,
-      "short_answer_loss": NaN,
-      "step": 399,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1599,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.9778195444082877e-05,
-      "long_answer_loss": 0.1599,
-      "loss": 0.1657,
-      "short_answer_loss": NaN,
-      "step": 400,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1777,
-      "grad_norm": 1.8828125,
-      "learning_rate": 1.9751520481233445e-05,
-      "long_answer_loss": 0.1777,
-      "loss": 0.1851,
-      "short_answer_loss": NaN,
-      "step": 401,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1591,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.9724795651189214e-05,
-      "long_answer_loss": 0.1591,
-      "loss": 0.1582,
-      "short_answer_loss": NaN,
-      "step": 402,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1798,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.969802113773127e-05,
-      "long_answer_loss": 0.1798,
-      "loss": 0.1707,
-      "short_answer_loss": NaN,
-      "step": 403,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1611,
-      "grad_norm": 1.75,
-      "learning_rate": 1.967119712498236e-05,
-      "long_answer_loss": 0.1611,
-      "loss": 0.1632,
-      "short_answer_loss": NaN,
-      "step": 404,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.65,
-      "full_loss": 0.1505,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.9644323797405633e-05,
-      "long_answer_loss": 0.1505,
-      "loss": 0.1616,
-      "short_answer_loss": NaN,
-      "step": 405,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.2129,
-      "grad_norm": 1.625,
-      "learning_rate": 1.961740133980336e-05,
-      "long_answer_loss": 0.2129,
-      "loss": 0.1712,
-      "short_answer_loss": NaN,
-      "step": 406,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1631,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.959042993731567e-05,
-      "long_answer_loss": 0.1631,
-      "loss": 0.1586,
-      "short_answer_loss": NaN,
-      "step": 407,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1718,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.956340977541927e-05,
-      "long_answer_loss": 0.1718,
-      "loss": 0.1687,
-      "short_answer_loss": NaN,
-      "step": 408,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1629,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.9536341039926186e-05,
-      "long_answer_loss": 0.1629,
-      "loss": 0.1702,
-      "short_answer_loss": NaN,
-      "step": 409,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1599,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.9509223916982472e-05,
-      "long_answer_loss": 0.1599,
-      "loss": 0.1605,
-      "short_answer_loss": NaN,
-      "step": 410,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.66,
-      "full_loss": 0.1586,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.9482058593066923e-05,
-      "long_answer_loss": 0.1586,
-      "loss": 0.171,
-      "short_answer_loss": NaN,
-      "step": 411,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1573,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.9454845254989818e-05,
-      "long_answer_loss": 0.1573,
-      "loss": 0.1675,
-      "short_answer_loss": NaN,
-      "step": 412,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.177,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.9427584089891598e-05,
-      "long_answer_loss": 0.177,
-      "loss": 0.1752,
-      "short_answer_loss": NaN,
-      "step": 413,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1527,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.9400275285241624e-05,
-      "long_answer_loss": 0.1527,
-      "loss": 0.1598,
-      "short_answer_loss": NaN,
-      "step": 414,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1577,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.9372919028836855e-05,
-      "long_answer_loss": 0.1577,
-      "loss": 0.1734,
-      "short_answer_loss": NaN,
-      "step": 415,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.183,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.9345515508800556e-05,
-      "long_answer_loss": 0.183,
-      "loss": 0.1676,
-      "short_answer_loss": NaN,
-      "step": 416,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.67,
-      "full_loss": 0.1698,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.931806491358102e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.1672,
-      "short_answer_loss": NaN,
-      "step": 417,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1583,
-      "grad_norm": 1.8828125,
-      "learning_rate": 1.929056743195028e-05,
-      "long_answer_loss": 0.1583,
-      "loss": 0.172,
-      "short_answer_loss": NaN,
-      "step": 418,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1488,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.9263023253002773e-05,
-      "long_answer_loss": 0.1488,
-      "loss": 0.1674,
-      "short_answer_loss": NaN,
-      "step": 419,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1781,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.9235432566154084e-05,
-      "long_answer_loss": 0.1781,
-      "loss": 0.1593,
-      "short_answer_loss": NaN,
-      "step": 420,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1786,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.9207795561139614e-05,
-      "long_answer_loss": 0.1786,
-      "loss": 0.1674,
-      "short_answer_loss": NaN,
-      "step": 421,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1826,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.9180112428013286e-05,
-      "long_answer_loss": 0.1826,
-      "loss": 0.1615,
-      "short_answer_loss": NaN,
-      "step": 422,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.68,
-      "full_loss": 0.1832,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.915238335714623e-05,
-      "long_answer_loss": 0.1832,
-      "loss": 0.1728,
-      "short_answer_loss": NaN,
-      "step": 423,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1843,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.9124608539225496e-05,
-      "long_answer_loss": 0.1843,
-      "loss": 0.1677,
-      "short_answer_loss": NaN,
-      "step": 424,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.169,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.909678816525271e-05,
-      "long_answer_loss": 0.169,
-      "loss": 0.1727,
-      "short_answer_loss": NaN,
-      "step": 425,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1327,
-      "grad_norm": 1.625,
-      "learning_rate": 1.9068922426542783e-05,
-      "long_answer_loss": 0.1327,
-      "loss": 0.1565,
-      "short_answer_loss": NaN,
-      "step": 426,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1685,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.9041011514722602e-05,
-      "long_answer_loss": 0.1685,
-      "loss": 0.1745,
-      "short_answer_loss": NaN,
-      "step": 427,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1876,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.901305562172968e-05,
-      "long_answer_loss": 0.1876,
-      "loss": 0.1722,
-      "short_answer_loss": NaN,
-      "step": 428,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.69,
-      "full_loss": 0.1595,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.898505493981087e-05,
-      "long_answer_loss": 0.1595,
-      "loss": 0.1667,
-      "short_answer_loss": NaN,
-      "step": 429,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1656,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.895700966152103e-05,
-      "long_answer_loss": 0.1656,
-      "loss": 0.1568,
-      "short_answer_loss": NaN,
-      "step": 430,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1723,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.8928919979721678e-05,
-      "long_answer_loss": 0.1723,
-      "loss": 0.1661,
-      "short_answer_loss": NaN,
-      "step": 431,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1612,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.8900786087579712e-05,
-      "long_answer_loss": 0.1612,
-      "loss": 0.1571,
-      "short_answer_loss": NaN,
-      "step": 432,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1496,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.8872608178566043e-05,
-      "long_answer_loss": 0.1496,
-      "loss": 0.1661,
-      "short_answer_loss": NaN,
-      "step": 433,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1647,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.8844386446454275e-05,
-      "long_answer_loss": 0.1647,
-      "loss": 0.1686,
-      "short_answer_loss": NaN,
-      "step": 434,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1544,
-      "grad_norm": 1.890625,
-      "learning_rate": 1.881612108531938e-05,
-      "long_answer_loss": 0.1544,
-      "loss": 0.1683,
-      "short_answer_loss": NaN,
-      "step": 435,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.7,
-      "full_loss": 0.1494,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.878781228953635e-05,
-      "long_answer_loss": 0.1494,
-      "loss": 0.1597,
-      "short_answer_loss": NaN,
-      "step": 436,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.171,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.8759460253778877e-05,
-      "long_answer_loss": 0.171,
-      "loss": 0.1691,
-      "short_answer_loss": NaN,
-      "step": 437,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1657,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.8731065173018e-05,
-      "long_answer_loss": 0.1657,
-      "loss": 0.1576,
-      "short_answer_loss": NaN,
-      "step": 438,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.17,
-      "grad_norm": 1.8515625,
-      "learning_rate": 1.870262724252077e-05,
-      "long_answer_loss": 0.17,
-      "loss": 0.1781,
-      "short_answer_loss": NaN,
-      "step": 439,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1849,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.8674146657848908e-05,
-      "long_answer_loss": 0.1849,
-      "loss": 0.1736,
-      "short_answer_loss": NaN,
-      "step": 440,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.1792,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.8645623614857455e-05,
-      "long_answer_loss": 0.1792,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 441,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.71,
-      "full_loss": 0.2047,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.8617058309693437e-05,
-      "long_answer_loss": 0.2047,
-      "loss": 0.1788,
-      "short_answer_loss": NaN,
-      "step": 442,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.163,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.85884509387945e-05,
-      "long_answer_loss": 0.163,
-      "loss": 0.1705,
-      "short_answer_loss": NaN,
-      "step": 443,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.1636,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.855980169888757e-05,
-      "long_answer_loss": 0.1636,
-      "loss": 0.1819,
-      "short_answer_loss": NaN,
-      "step": 444,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.1581,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.85311107869875e-05,
-      "long_answer_loss": 0.1581,
-      "loss": 0.1666,
-      "short_answer_loss": NaN,
-      "step": 445,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.1779,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.850237840039571e-05,
-      "long_answer_loss": 0.1779,
-      "loss": 0.1726,
-      "short_answer_loss": NaN,
-      "step": 446,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.1499,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.8473604736698835e-05,
-      "long_answer_loss": 0.1499,
-      "loss": 0.1626,
-      "short_answer_loss": NaN,
-      "step": 447,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.72,
-      "full_loss": 0.171,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.844478999376736e-05,
-      "long_answer_loss": 0.171,
-      "loss": 0.1553,
-      "short_answer_loss": NaN,
-      "step": 448,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1674,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.841593436975427e-05,
-      "long_answer_loss": 0.1674,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 449,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1619,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.838703806309367e-05,
-      "long_answer_loss": 0.1619,
-      "loss": 0.163,
-      "short_answer_loss": NaN,
-      "step": 450,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1672,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.8358101272499443e-05,
-      "long_answer_loss": 0.1672,
-      "loss": 0.1597,
-      "short_answer_loss": NaN,
-      "step": 451,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1467,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.8329124196963864e-05,
-      "long_answer_loss": 0.1467,
-      "loss": 0.1748,
-      "short_answer_loss": NaN,
-      "step": 452,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1644,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.830010703575624e-05,
-      "long_answer_loss": 0.1644,
-      "loss": 0.1643,
-      "short_answer_loss": NaN,
-      "step": 453,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.73,
-      "full_loss": 0.1672,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.827104998842154e-05,
-      "long_answer_loss": 0.1672,
-      "loss": 0.1646,
-      "short_answer_loss": NaN,
-      "step": 454,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1646,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.8241953254779027e-05,
-      "long_answer_loss": 0.1646,
-      "loss": 0.1643,
-      "short_answer_loss": NaN,
-      "step": 455,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1869,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.8212817034920864e-05,
-      "long_answer_loss": 0.1869,
-      "loss": 0.1684,
-      "short_answer_loss": NaN,
-      "step": 456,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1618,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.818364152921077e-05,
-      "long_answer_loss": 0.1618,
-      "loss": 0.1577,
-      "short_answer_loss": NaN,
-      "step": 457,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1658,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.8154426938282615e-05,
-      "long_answer_loss": 0.1658,
-      "loss": 0.1688,
-      "short_answer_loss": NaN,
-      "step": 458,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1681,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.8125173463039048e-05,
-      "long_answer_loss": 0.1681,
-      "loss": 0.1663,
-      "short_answer_loss": NaN,
-      "step": 459,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.74,
-      "full_loss": 0.1868,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.8095881304650123e-05,
-      "long_answer_loss": 0.1868,
-      "loss": 0.1678,
-      "short_answer_loss": NaN,
-      "step": 460,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1519,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.8066550664551904e-05,
-      "long_answer_loss": 0.1519,
-      "loss": 0.1604,
-      "short_answer_loss": NaN,
-      "step": 461,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1849,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.8037181744445093e-05,
-      "long_answer_loss": 0.1849,
-      "loss": 0.1701,
-      "short_answer_loss": NaN,
-      "step": 462,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1386,
-      "grad_norm": 1.625,
-      "learning_rate": 1.8007774746293628e-05,
-      "long_answer_loss": 0.1386,
-      "loss": 0.1604,
-      "short_answer_loss": NaN,
-      "step": 463,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1884,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.7978329872323308e-05,
-      "long_answer_loss": 0.1884,
-      "loss": 0.16,
-      "short_answer_loss": NaN,
-      "step": 464,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1752,
-      "grad_norm": 1.8671875,
-      "learning_rate": 1.7948847325020394e-05,
-      "long_answer_loss": 0.1752,
-      "loss": 0.1646,
-      "short_answer_loss": NaN,
-      "step": 465,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.75,
-      "full_loss": 0.1523,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.7919327307130217e-05,
-      "long_answer_loss": 0.1523,
-      "loss": 0.1599,
-      "short_answer_loss": NaN,
-      "step": 466,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1735,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.7889770021655787e-05,
-      "long_answer_loss": 0.1735,
-      "loss": 0.1826,
-      "short_answer_loss": NaN,
-      "step": 467,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.16,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.78601756718564e-05,
-      "long_answer_loss": 0.16,
-      "loss": 0.1615,
-      "short_answer_loss": NaN,
-      "step": 468,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1622,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.783054446124622e-05,
-      "long_answer_loss": 0.1622,
-      "loss": 0.1562,
-      "short_answer_loss": NaN,
-      "step": 469,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1735,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.7800876593592912e-05,
-      "long_answer_loss": 0.1735,
-      "loss": 0.1663,
-      "short_answer_loss": NaN,
-      "step": 470,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.148,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.777117227291622e-05,
-      "long_answer_loss": 0.148,
-      "loss": 0.167,
-      "short_answer_loss": NaN,
-      "step": 471,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1756,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.7741431703486562e-05,
-      "long_answer_loss": 0.1756,
-      "loss": 0.1616,
-      "short_answer_loss": NaN,
-      "step": 472,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.76,
-      "full_loss": 0.1404,
-      "grad_norm": 1.75,
-      "learning_rate": 1.7711655089823638e-05,
-      "long_answer_loss": 0.1404,
-      "loss": 0.1633,
-      "short_answer_loss": NaN,
-      "step": 473,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1554,
-      "grad_norm": 1.859375,
-      "learning_rate": 1.7681842636695007e-05,
-      "long_answer_loss": 0.1554,
-      "loss": 0.1644,
-      "short_answer_loss": NaN,
-      "step": 474,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1773,
-      "grad_norm": 1.859375,
-      "learning_rate": 1.7651994549114702e-05,
-      "long_answer_loss": 0.1773,
-      "loss": 0.1651,
-      "short_answer_loss": NaN,
-      "step": 475,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1536,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.7622111032341797e-05,
-      "long_answer_loss": 0.1536,
-      "loss": 0.1628,
-      "short_answer_loss": NaN,
-      "step": 476,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1713,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.7592192291879008e-05,
-      "long_answer_loss": 0.1713,
-      "loss": 0.1678,
-      "short_answer_loss": NaN,
-      "step": 477,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1917,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.756223853347127e-05,
-      "long_answer_loss": 0.1917,
-      "loss": 0.1679,
-      "short_answer_loss": NaN,
-      "step": 478,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.77,
-      "full_loss": 0.1724,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.7532249963104344e-05,
-      "long_answer_loss": 0.1724,
-      "loss": 0.167,
-      "short_answer_loss": NaN,
-      "step": 479,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1845,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.7502226787003378e-05,
-      "long_answer_loss": 0.1845,
-      "loss": 0.1716,
-      "short_answer_loss": NaN,
-      "step": 480,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1741,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.747216921163149e-05,
-      "long_answer_loss": 0.1741,
-      "loss": 0.1599,
-      "short_answer_loss": NaN,
-      "step": 481,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1524,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.7442077443688364e-05,
-      "long_answer_loss": 0.1524,
-      "loss": 0.1597,
-      "short_answer_loss": NaN,
-      "step": 482,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1621,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.741195169010882e-05,
-      "long_answer_loss": 0.1621,
-      "loss": 0.1614,
-      "short_answer_loss": NaN,
-      "step": 483,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1449,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.7381792158061378e-05,
-      "long_answer_loss": 0.1449,
-      "loss": 0.1676,
-      "short_answer_loss": NaN,
-      "step": 484,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.78,
-      "full_loss": 0.1858,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.7351599054946853e-05,
-      "long_answer_loss": 0.1858,
-      "loss": 0.1664,
-      "short_answer_loss": NaN,
-      "step": 485,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1459,
-      "grad_norm": 1.625,
-      "learning_rate": 1.732137258839693e-05,
-      "long_answer_loss": 0.1459,
-      "loss": 0.1635,
-      "short_answer_loss": NaN,
-      "step": 486,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1591,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.7291112966272707e-05,
-      "long_answer_loss": 0.1591,
-      "loss": 0.1481,
-      "short_answer_loss": NaN,
-      "step": 487,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1824,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.7260820396663307e-05,
-      "long_answer_loss": 0.1824,
-      "loss": 0.1691,
-      "short_answer_loss": NaN,
-      "step": 488,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.2103,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.723049508788442e-05,
-      "long_answer_loss": 0.2103,
-      "loss": 0.1607,
-      "short_answer_loss": NaN,
-      "step": 489,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1865,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.720013724847686e-05,
-      "long_answer_loss": 0.1865,
-      "loss": 0.1544,
-      "short_answer_loss": NaN,
-      "step": 490,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.79,
-      "full_loss": 0.1951,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.716974708720517e-05,
-      "long_answer_loss": 0.1951,
-      "loss": 0.1694,
-      "short_answer_loss": NaN,
-      "step": 491,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1457,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.7139324813056155e-05,
-      "long_answer_loss": 0.1457,
-      "loss": 0.1615,
-      "short_answer_loss": NaN,
-      "step": 492,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1946,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.7108870635237444e-05,
-      "long_answer_loss": 0.1946,
-      "loss": 0.1607,
-      "short_answer_loss": NaN,
-      "step": 493,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1513,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.7078384763176084e-05,
-      "long_answer_loss": 0.1513,
-      "loss": 0.1649,
-      "short_answer_loss": NaN,
-      "step": 494,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1478,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.7047867406517047e-05,
-      "long_answer_loss": 0.1478,
-      "loss": 0.1599,
-      "short_answer_loss": NaN,
-      "step": 495,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1656,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.7017318775121845e-05,
-      "long_answer_loss": 0.1656,
-      "loss": 0.1581,
-      "short_answer_loss": NaN,
-      "step": 496,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.8,
-      "full_loss": 0.1545,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.6986739079067047e-05,
-      "long_answer_loss": 0.1545,
-      "loss": 0.1649,
-      "short_answer_loss": NaN,
-      "step": 497,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.155,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.6956128528642842e-05,
-      "long_answer_loss": 0.155,
-      "loss": 0.1563,
-      "short_answer_loss": NaN,
-      "step": 498,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1534,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.6925487334351613e-05,
-      "long_answer_loss": 0.1534,
-      "loss": 0.1636,
-      "short_answer_loss": NaN,
-      "step": 499,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.169,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.6894815706906458e-05,
-      "long_answer_loss": 0.169,
-      "loss": 0.1567,
-      "short_answer_loss": NaN,
-      "step": 500,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.168,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.686411385722977e-05,
-      "long_answer_loss": 0.168,
-      "loss": 0.16,
-      "short_answer_loss": NaN,
-      "step": 501,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1441,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.683338199645177e-05,
-      "long_answer_loss": 0.1441,
-      "loss": 0.1669,
-      "short_answer_loss": NaN,
-      "step": 502,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1572,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.6802620335909054e-05,
-      "long_answer_loss": 0.1572,
-      "loss": 0.1643,
-      "short_answer_loss": NaN,
-      "step": 503,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.81,
-      "full_loss": 0.1503,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.6771829087143156e-05,
-      "long_answer_loss": 0.1503,
-      "loss": 0.1601,
-      "short_answer_loss": NaN,
-      "step": 504,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1735,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.6741008461899073e-05,
-      "long_answer_loss": 0.1735,
-      "loss": 0.1561,
-      "short_answer_loss": NaN,
-      "step": 505,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.183,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.6710158672123818e-05,
-      "long_answer_loss": 0.183,
-      "loss": 0.1685,
-      "short_answer_loss": NaN,
-      "step": 506,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.16,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.6679279929964968e-05,
-      "long_answer_loss": 0.16,
-      "loss": 0.1582,
-      "short_answer_loss": NaN,
-      "step": 507,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1683,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.6648372447769197e-05,
-      "long_answer_loss": 0.1683,
-      "loss": 0.167,
-      "short_answer_loss": NaN,
-      "step": 508,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1724,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.6617436438080812e-05,
-      "long_answer_loss": 0.1724,
-      "loss": 0.16,
-      "short_answer_loss": NaN,
-      "step": 509,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.82,
-      "full_loss": 0.1824,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.6586472113640306e-05,
-      "long_answer_loss": 0.1824,
-      "loss": 0.1644,
-      "short_answer_loss": NaN,
-      "step": 510,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1504,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.6555479687382887e-05,
-      "long_answer_loss": 0.1504,
-      "loss": 0.1576,
-      "short_answer_loss": NaN,
-      "step": 511,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1488,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.6524459372437004e-05,
-      "long_answer_loss": 0.1488,
-      "loss": 0.1612,
-      "short_answer_loss": NaN,
-      "step": 512,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1622,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.64934113821229e-05,
-      "long_answer_loss": 0.1622,
-      "loss": 0.1553,
-      "short_answer_loss": NaN,
-      "step": 513,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1501,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.6462335929951133e-05,
-      "long_answer_loss": 0.1501,
-      "loss": 0.1545,
-      "short_answer_loss": NaN,
-      "step": 514,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1467,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.643123322962111e-05,
-      "long_answer_loss": 0.1467,
-      "loss": 0.154,
-      "short_answer_loss": NaN,
-      "step": 515,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.83,
-      "full_loss": 0.1546,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.6400103495019618e-05,
-      "long_answer_loss": 0.1546,
-      "loss": 0.1612,
-      "short_answer_loss": NaN,
-      "step": 516,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1756,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.6368946940219352e-05,
-      "long_answer_loss": 0.1756,
-      "loss": 0.1622,
-      "short_answer_loss": NaN,
-      "step": 517,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1914,
-      "grad_norm": 1.75,
-      "learning_rate": 1.633776377947745e-05,
-      "long_answer_loss": 0.1914,
-      "loss": 0.1644,
-      "short_answer_loss": NaN,
-      "step": 518,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1482,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.6306554227233994e-05,
-      "long_answer_loss": 0.1482,
-      "loss": 0.1614,
-      "short_answer_loss": NaN,
-      "step": 519,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.147,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.6275318498110585e-05,
-      "long_answer_loss": 0.147,
-      "loss": 0.1617,
-      "short_answer_loss": NaN,
-      "step": 520,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1683,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.6244056806908816e-05,
-      "long_answer_loss": 0.1683,
-      "loss": 0.1655,
-      "short_answer_loss": NaN,
-      "step": 521,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.84,
-      "full_loss": 0.1623,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.621276936860882e-05,
-      "long_answer_loss": 0.1623,
-      "loss": 0.1619,
-      "short_answer_loss": NaN,
-      "step": 522,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1877,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.6181456398367788e-05,
-      "long_answer_loss": 0.1877,
-      "loss": 0.1621,
-      "short_answer_loss": NaN,
-      "step": 523,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1704,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.6150118111518493e-05,
-      "long_answer_loss": 0.1704,
-      "loss": 0.1613,
-      "short_answer_loss": NaN,
-      "step": 524,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1437,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.6118754723567798e-05,
-      "long_answer_loss": 0.1437,
-      "loss": 0.1453,
-      "short_answer_loss": NaN,
-      "step": 525,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1142,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.608736645019518e-05,
-      "long_answer_loss": 0.1142,
-      "loss": 0.1564,
-      "short_answer_loss": NaN,
-      "step": 526,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.1551,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.605595350725126e-05,
-      "long_answer_loss": 0.1551,
-      "loss": 0.1566,
-      "short_answer_loss": NaN,
-      "step": 527,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.85,
-      "full_loss": 0.2099,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.6024516110756296e-05,
-      "long_answer_loss": 0.2099,
-      "loss": 0.1686,
-      "short_answer_loss": NaN,
-      "step": 528,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1506,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.5993054476898708e-05,
-      "long_answer_loss": 0.1506,
-      "loss": 0.1566,
-      "short_answer_loss": NaN,
-      "step": 529,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1487,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.59615688220336e-05,
-      "long_answer_loss": 0.1487,
-      "loss": 0.1565,
-      "short_answer_loss": NaN,
-      "step": 530,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1496,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.593005936268125e-05,
-      "long_answer_loss": 0.1496,
-      "loss": 0.1567,
-      "short_answer_loss": NaN,
-      "step": 531,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1571,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.5898526315525646e-05,
-      "long_answer_loss": 0.1571,
-      "loss": 0.1597,
-      "short_answer_loss": NaN,
-      "step": 532,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1515,
-      "grad_norm": 1.8515625,
-      "learning_rate": 1.5866969897412984e-05,
-      "long_answer_loss": 0.1515,
-      "loss": 0.1731,
-      "short_answer_loss": NaN,
-      "step": 533,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.1773,
-      "grad_norm": 1.7421875,
-      "learning_rate": 1.583539032535017e-05,
-      "long_answer_loss": 0.1773,
-      "loss": 0.1649,
-      "short_answer_loss": NaN,
-      "step": 534,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.86,
-      "full_loss": 0.16,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.5803787816503336e-05,
-      "long_answer_loss": 0.16,
-      "loss": 0.1577,
-      "short_answer_loss": NaN,
-      "step": 535,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1701,
-      "grad_norm": 1.875,
-      "learning_rate": 1.577216258819635e-05,
-      "long_answer_loss": 0.1701,
-      "loss": 0.1614,
-      "short_answer_loss": NaN,
-      "step": 536,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.137,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.5740514857909312e-05,
-      "long_answer_loss": 0.137,
-      "loss": 0.1698,
-      "short_answer_loss": NaN,
-      "step": 537,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1651,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.570884484327707e-05,
-      "long_answer_loss": 0.1651,
-      "loss": 0.1594,
-      "short_answer_loss": NaN,
-      "step": 538,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.132,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.5677152762087714e-05,
-      "long_answer_loss": 0.132,
-      "loss": 0.1594,
-      "short_answer_loss": NaN,
-      "step": 539,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1649,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.5645438832281077e-05,
-      "long_answer_loss": 0.1649,
-      "loss": 0.1537,
-      "short_answer_loss": NaN,
-      "step": 540,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.87,
-      "full_loss": 0.1636,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.561370327194725e-05,
-      "long_answer_loss": 0.1636,
-      "loss": 0.1579,
-      "short_answer_loss": NaN,
-      "step": 541,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1688,
-      "grad_norm": 1.75,
-      "learning_rate": 1.558194629932506e-05,
-      "long_answer_loss": 0.1688,
-      "loss": 0.1685,
-      "short_answer_loss": NaN,
-      "step": 542,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1594,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.5550168132800585e-05,
-      "long_answer_loss": 0.1594,
-      "loss": 0.1544,
-      "short_answer_loss": NaN,
-      "step": 543,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1528,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.5518368990905664e-05,
-      "long_answer_loss": 0.1528,
-      "loss": 0.1653,
-      "short_answer_loss": NaN,
-      "step": 544,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1369,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.5486549092316355e-05,
-      "long_answer_loss": 0.1369,
-      "loss": 0.1556,
-      "short_answer_loss": NaN,
-      "step": 545,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.1606,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.545470865585147e-05,
-      "long_answer_loss": 0.1606,
-      "loss": 0.1528,
-      "short_answer_loss": NaN,
-      "step": 546,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.88,
-      "full_loss": 0.171,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.5422847900471063e-05,
-      "long_answer_loss": 0.171,
-      "loss": 0.1659,
-      "short_answer_loss": NaN,
-      "step": 547,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1376,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.53909670452749e-05,
-      "long_answer_loss": 0.1376,
-      "loss": 0.1562,
-      "short_answer_loss": NaN,
-      "step": 548,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.161,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.5359066309500974e-05,
-      "long_answer_loss": 0.161,
-      "loss": 0.1572,
-      "short_answer_loss": NaN,
-      "step": 549,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1466,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.5327145912524e-05,
-      "long_answer_loss": 0.1466,
-      "loss": 0.1558,
-      "short_answer_loss": NaN,
-      "step": 550,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1694,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.5295206073853896e-05,
-      "long_answer_loss": 0.1694,
-      "loss": 0.1675,
-      "short_answer_loss": NaN,
-      "step": 551,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1796,
-      "grad_norm": 1.84375,
-      "learning_rate": 1.526324701313427e-05,
-      "long_answer_loss": 0.1796,
-      "loss": 0.16,
-      "short_answer_loss": NaN,
-      "step": 552,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.89,
-      "full_loss": 0.1545,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.5231268950140926e-05,
-      "long_answer_loss": 0.1545,
-      "loss": 0.1545,
-      "short_answer_loss": NaN,
-      "step": 553,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.147,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.5199272104780332e-05,
-      "long_answer_loss": 0.147,
-      "loss": 0.1467,
-      "short_answer_loss": NaN,
-      "step": 554,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.149,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.5167256697088128e-05,
-      "long_answer_loss": 0.149,
-      "loss": 0.1609,
-      "short_answer_loss": NaN,
-      "step": 555,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.165,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.5135222947227598e-05,
-      "long_answer_loss": 0.165,
-      "loss": 0.158,
-      "short_answer_loss": NaN,
-      "step": 556,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1821,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.510317107548816e-05,
-      "long_answer_loss": 0.1821,
-      "loss": 0.1679,
-      "short_answer_loss": NaN,
-      "step": 557,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1337,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.507110130228386e-05,
-      "long_answer_loss": 0.1337,
-      "loss": 0.1562,
-      "short_answer_loss": NaN,
-      "step": 558,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.9,
-      "full_loss": 0.1507,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.5039013848151839e-05,
-      "long_answer_loss": 0.1507,
-      "loss": 0.1464,
-      "short_answer_loss": NaN,
-      "step": 559,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1295,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.5006908933750829e-05,
-      "long_answer_loss": 0.1295,
-      "loss": 0.1484,
-      "short_answer_loss": NaN,
-      "step": 560,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1581,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.4974786779859642e-05,
-      "long_answer_loss": 0.1581,
-      "loss": 0.1511,
-      "short_answer_loss": NaN,
-      "step": 561,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1504,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.4942647607375629e-05,
-      "long_answer_loss": 0.1504,
-      "loss": 0.1664,
-      "short_answer_loss": NaN,
-      "step": 562,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1607,
-      "grad_norm": 1.8046875,
-      "learning_rate": 1.4910491637313176e-05,
-      "long_answer_loss": 0.1607,
-      "loss": 0.1653,
-      "short_answer_loss": NaN,
-      "step": 563,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.168,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.4878319090802196e-05,
-      "long_answer_loss": 0.168,
-      "loss": 0.1615,
-      "short_answer_loss": NaN,
-      "step": 564,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.91,
-      "full_loss": 0.1271,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.4846130189086577e-05,
-      "long_answer_loss": 0.1271,
-      "loss": 0.1421,
-      "short_answer_loss": NaN,
-      "step": 565,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1694,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.4813925153522693e-05,
-      "long_answer_loss": 0.1694,
-      "loss": 0.1608,
-      "short_answer_loss": NaN,
-      "step": 566,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1439,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.4781704205577856e-05,
-      "long_answer_loss": 0.1439,
-      "loss": 0.1462,
-      "short_answer_loss": NaN,
-      "step": 567,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1338,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.4749467566828808e-05,
-      "long_answer_loss": 0.1338,
-      "loss": 0.153,
-      "short_answer_loss": NaN,
-      "step": 568,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.169,
-      "grad_norm": 1.7890625,
-      "learning_rate": 1.4717215458960198e-05,
-      "long_answer_loss": 0.169,
-      "loss": 0.1626,
-      "short_answer_loss": NaN,
-      "step": 569,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1542,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.4684948103763046e-05,
-      "long_answer_loss": 0.1542,
-      "loss": 0.1596,
-      "short_answer_loss": NaN,
-      "step": 570,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1458,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.465266572313323e-05,
-      "long_answer_loss": 0.1458,
-      "loss": 0.1556,
-      "short_answer_loss": NaN,
-      "step": 571,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.92,
-      "full_loss": 0.1489,
-      "grad_norm": 1.828125,
-      "learning_rate": 1.462036853906995e-05,
-      "long_answer_loss": 0.1489,
-      "loss": 0.1652,
-      "short_answer_loss": NaN,
-      "step": 572,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1884,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.4588056773674209e-05,
-      "long_answer_loss": 0.1884,
-      "loss": 0.164,
-      "short_answer_loss": NaN,
-      "step": 573,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1575,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.4555730649147285e-05,
-      "long_answer_loss": 0.1575,
-      "loss": 0.1548,
-      "short_answer_loss": NaN,
-      "step": 574,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1655,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.4523390387789193e-05,
-      "long_answer_loss": 0.1655,
-      "loss": 0.1563,
-      "short_answer_loss": NaN,
-      "step": 575,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1531,
-      "grad_norm": 1.859375,
-      "learning_rate": 1.4491036211997175e-05,
-      "long_answer_loss": 0.1531,
-      "loss": 0.1585,
-      "short_answer_loss": NaN,
-      "step": 576,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1501,
-      "grad_norm": 1.875,
-      "learning_rate": 1.4458668344264151e-05,
-      "long_answer_loss": 0.1501,
-      "loss": 0.1459,
-      "short_answer_loss": NaN,
-      "step": 577,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.93,
-      "full_loss": 0.1541,
-      "grad_norm": 1.75,
-      "learning_rate": 1.4426287007177197e-05,
-      "long_answer_loss": 0.1541,
-      "loss": 0.1511,
-      "short_answer_loss": NaN,
-      "step": 578,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1538,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.4393892423416025e-05,
-      "long_answer_loss": 0.1538,
-      "loss": 0.1514,
-      "short_answer_loss": NaN,
-      "step": 579,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1665,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.4361484815751434e-05,
-      "long_answer_loss": 0.1665,
-      "loss": 0.1508,
-      "short_answer_loss": NaN,
-      "step": 580,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1698,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.432906440704378e-05,
-      "long_answer_loss": 0.1698,
-      "loss": 0.155,
-      "short_answer_loss": NaN,
-      "step": 581,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1477,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.4296631420241463e-05,
-      "long_answer_loss": 0.1477,
-      "loss": 0.1553,
-      "short_answer_loss": NaN,
-      "step": 582,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1408,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.4264186078379369e-05,
-      "long_answer_loss": 0.1408,
-      "loss": 0.1527,
-      "short_answer_loss": NaN,
-      "step": 583,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.94,
-      "full_loss": 0.1557,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.4231728604577352e-05,
-      "long_answer_loss": 0.1557,
-      "loss": 0.1566,
-      "short_answer_loss": NaN,
-      "step": 584,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1524,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.4199259222038694e-05,
-      "long_answer_loss": 0.1524,
-      "loss": 0.1529,
-      "short_answer_loss": NaN,
-      "step": 585,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1409,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.416677815404857e-05,
-      "long_answer_loss": 0.1409,
-      "loss": 0.1479,
-      "short_answer_loss": NaN,
-      "step": 586,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1292,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.4134285623972514e-05,
-      "long_answer_loss": 0.1292,
-      "loss": 0.1496,
-      "short_answer_loss": NaN,
-      "step": 587,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1612,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.4101781855254883e-05,
-      "long_answer_loss": 0.1612,
-      "loss": 0.1509,
-      "short_answer_loss": NaN,
-      "step": 588,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1768,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.406926707141732e-05,
-      "long_answer_loss": 0.1768,
-      "loss": 0.1594,
-      "short_answer_loss": NaN,
-      "step": 589,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.95,
-      "full_loss": 0.1408,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.4036741496057213e-05,
-      "long_answer_loss": 0.1408,
-      "loss": 0.1453,
-      "short_answer_loss": NaN,
-      "step": 590,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1442,
-      "grad_norm": 1.7734375,
-      "learning_rate": 1.4004205352846164e-05,
-      "long_answer_loss": 0.1442,
-      "loss": 0.1534,
-      "short_answer_loss": NaN,
-      "step": 591,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1261,
-      "grad_norm": 1.78125,
-      "learning_rate": 1.3971658865528451e-05,
-      "long_answer_loss": 0.1261,
-      "loss": 0.1595,
-      "short_answer_loss": NaN,
-      "step": 592,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.182,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.3939102257919481e-05,
-      "long_answer_loss": 0.182,
-      "loss": 0.1481,
-      "short_answer_loss": NaN,
-      "step": 593,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1522,
-      "grad_norm": 1.75,
-      "learning_rate": 1.390653575390426e-05,
-      "long_answer_loss": 0.1522,
-      "loss": 0.1629,
-      "short_answer_loss": NaN,
-      "step": 594,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1329,
-      "grad_norm": 1.8125,
-      "learning_rate": 1.3873959577435847e-05,
-      "long_answer_loss": 0.1329,
-      "loss": 0.1608,
-      "short_answer_loss": NaN,
-      "step": 595,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.96,
-      "full_loss": 0.1544,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.3841373952533812e-05,
-      "long_answer_loss": 0.1544,
-      "loss": 0.1516,
-      "short_answer_loss": NaN,
-      "step": 596,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1549,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.3808779103282712e-05,
-      "long_answer_loss": 0.1549,
-      "loss": 0.1498,
-      "short_answer_loss": NaN,
-      "step": 597,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1468,
-      "grad_norm": 1.625,
-      "learning_rate": 1.3776175253830531e-05,
-      "long_answer_loss": 0.1468,
-      "loss": 0.1496,
-      "short_answer_loss": NaN,
-      "step": 598,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1515,
-      "grad_norm": 1.703125,
-      "learning_rate": 1.3743562628387141e-05,
-      "long_answer_loss": 0.1515,
-      "loss": 0.1535,
-      "short_answer_loss": NaN,
-      "step": 599,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1576,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.3710941451222776e-05,
-      "long_answer_loss": 0.1576,
-      "loss": 0.1506,
-      "short_answer_loss": NaN,
-      "step": 600,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1565,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.367831194666646e-05,
-      "long_answer_loss": 0.1565,
-      "loss": 0.1484,
-      "short_answer_loss": NaN,
-      "step": 601,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1385,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.3645674339104508e-05,
-      "long_answer_loss": 0.1385,
-      "loss": 0.1581,
-      "short_answer_loss": NaN,
-      "step": 602,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.97,
-      "full_loss": 0.1386,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.3613028852978934e-05,
-      "long_answer_loss": 0.1386,
-      "loss": 0.1505,
-      "short_answer_loss": NaN,
-      "step": 603,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1686,
-      "grad_norm": 1.625,
-      "learning_rate": 1.3580375712785945e-05,
-      "long_answer_loss": 0.1686,
-      "loss": 0.1552,
-      "short_answer_loss": NaN,
-      "step": 604,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1538,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.354771514307438e-05,
-      "long_answer_loss": 0.1538,
-      "loss": 0.1548,
-      "short_answer_loss": NaN,
-      "step": 605,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1412,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.3515047368444169e-05,
-      "long_answer_loss": 0.1412,
-      "loss": 0.1488,
-      "short_answer_loss": NaN,
-      "step": 606,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1467,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.3482372613544788e-05,
-      "long_answer_loss": 0.1467,
-      "loss": 0.15,
-      "short_answer_loss": NaN,
-      "step": 607,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.191,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.3449691103073714e-05,
-      "long_answer_loss": 0.191,
-      "loss": 0.1575,
-      "short_answer_loss": NaN,
-      "step": 608,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.98,
-      "full_loss": 0.1561,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.3417003061774886e-05,
-      "long_answer_loss": 0.1561,
-      "loss": 0.1569,
-      "short_answer_loss": NaN,
-      "step": 609,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1616,
-      "grad_norm": 1.6875,
-      "learning_rate": 1.3384308714437146e-05,
-      "long_answer_loss": 0.1616,
-      "loss": 0.1667,
-      "short_answer_loss": NaN,
-      "step": 610,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1367,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.3351608285892708e-05,
-      "long_answer_loss": 0.1367,
-      "loss": 0.1531,
-      "short_answer_loss": NaN,
-      "step": 611,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1453,
-      "grad_norm": 1.71875,
-      "learning_rate": 1.3318902001015602e-05,
-      "long_answer_loss": 0.1453,
-      "loss": 0.1489,
-      "short_answer_loss": NaN,
-      "step": 612,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1351,
-      "grad_norm": 1.7265625,
-      "learning_rate": 1.328619008472013e-05,
-      "long_answer_loss": 0.1351,
-      "loss": 0.1525,
-      "short_answer_loss": NaN,
-      "step": 613,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1462,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.3253472761959326e-05,
-      "long_answer_loss": 0.1462,
-      "loss": 0.1559,
-      "short_answer_loss": NaN,
-      "step": 614,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 0.99,
-      "full_loss": 0.1419,
-      "grad_norm": 1.7578125,
-      "learning_rate": 1.3220750257723397e-05,
-      "long_answer_loss": 0.1419,
-      "loss": 0.159,
-      "short_answer_loss": NaN,
-      "step": 615,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1622,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.3188022797038183e-05,
-      "long_answer_loss": 0.1622,
-      "loss": 0.1514,
-      "short_answer_loss": NaN,
-      "step": 616,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1706,
-      "grad_norm": 1.75,
-      "learning_rate": 1.3155290604963613e-05,
-      "long_answer_loss": 0.1706,
-      "loss": 0.1585,
-      "short_answer_loss": NaN,
-      "step": 617,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.1639,
-      "grad_norm": 1.765625,
-      "learning_rate": 1.3122553906592142e-05,
-      "long_answer_loss": 0.1639,
-      "loss": 0.1556,
-      "short_answer_loss": NaN,
-      "step": 618,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.114,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.3089812927047224e-05,
-      "long_answer_loss": 0.114,
-      "loss": 0.1224,
-      "short_answer_loss": NaN,
-      "step": 619,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.0845,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.3057067891481752e-05,
-      "long_answer_loss": 0.0845,
-      "loss": 0.0856,
-      "short_answer_loss": NaN,
-      "step": 620,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.0,
-      "full_loss": 0.0776,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.3024319025076509e-05,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 621,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0675,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.2991566553038623e-05,
-      "long_answer_loss": 0.0675,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 622,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0837,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.2958810700600017e-05,
-      "long_answer_loss": 0.0837,
-      "loss": 0.0857,
-      "short_answer_loss": NaN,
-      "step": 623,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0696,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.2926051693015858e-05,
-      "long_answer_loss": 0.0696,
-      "loss": 0.0811,
-      "short_answer_loss": NaN,
-      "step": 624,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0957,
-      "grad_norm": 1.796875,
-      "learning_rate": 1.2893289755563017e-05,
-      "long_answer_loss": 0.0957,
-      "loss": 0.0848,
-      "short_answer_loss": NaN,
-      "step": 625,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0815,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.2860525113538505e-05,
-      "long_answer_loss": 0.0815,
-      "loss": 0.0819,
-      "short_answer_loss": NaN,
-      "step": 626,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.01,
-      "full_loss": 0.0937,
-      "grad_norm": 1.8203125,
-      "learning_rate": 1.2827757992257939e-05,
-      "long_answer_loss": 0.0937,
-      "loss": 0.0895,
-      "short_answer_loss": NaN,
-      "step": 627,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0934,
-      "grad_norm": 1.8359375,
-      "learning_rate": 1.2794988617053979e-05,
-      "long_answer_loss": 0.0934,
-      "loss": 0.082,
-      "short_answer_loss": NaN,
-      "step": 628,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0839,
-      "grad_norm": 1.8515625,
-      "learning_rate": 1.2762217213274788e-05,
-      "long_answer_loss": 0.0839,
-      "loss": 0.0818,
-      "short_answer_loss": NaN,
-      "step": 629,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0908,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.2729444006282481e-05,
-      "long_answer_loss": 0.0908,
-      "loss": 0.0785,
-      "short_answer_loss": NaN,
-      "step": 630,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0839,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.269666922145157e-05,
-      "long_answer_loss": 0.0839,
-      "loss": 0.0829,
-      "short_answer_loss": NaN,
-      "step": 631,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0815,
-      "grad_norm": 1.9140625,
-      "learning_rate": 1.266389308416742e-05,
-      "long_answer_loss": 0.0815,
-      "loss": 0.0868,
-      "short_answer_loss": NaN,
-      "step": 632,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.02,
-      "full_loss": 0.0666,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.2631115819824688e-05,
-      "long_answer_loss": 0.0666,
-      "loss": 0.0788,
-      "short_answer_loss": NaN,
-      "step": 633,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0738,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.2598337653825798e-05,
-      "long_answer_loss": 0.0738,
-      "loss": 0.083,
-      "short_answer_loss": NaN,
-      "step": 634,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.08,
-      "grad_norm": 1.7109375,
-      "learning_rate": 1.2565558811579359e-05,
-      "long_answer_loss": 0.08,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 635,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0682,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.2532779518498639e-05,
-      "long_answer_loss": 0.0682,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 636,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0862,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.25e-05,
-      "long_answer_loss": 0.0862,
-      "loss": 0.08,
-      "short_answer_loss": NaN,
-      "step": 637,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0853,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.2467220481501365e-05,
-      "long_answer_loss": 0.0853,
-      "loss": 0.0784,
-      "short_answer_loss": NaN,
-      "step": 638,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.0645,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.2434441188420644e-05,
-      "long_answer_loss": 0.0645,
-      "loss": 0.0767,
-      "short_answer_loss": NaN,
-      "step": 639,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.03,
-      "full_loss": 0.1023,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.2401662346174206e-05,
-      "long_answer_loss": 0.1023,
-      "loss": 0.0801,
-      "short_answer_loss": NaN,
-      "step": 640,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0786,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2368884180175313e-05,
-      "long_answer_loss": 0.0786,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 641,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0839,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.2336106915832585e-05,
-      "long_answer_loss": 0.0839,
-      "loss": 0.0813,
-      "short_answer_loss": NaN,
-      "step": 642,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0719,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.2303330778548433e-05,
-      "long_answer_loss": 0.0719,
-      "loss": 0.0821,
-      "short_answer_loss": NaN,
-      "step": 643,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0633,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2270555993717521e-05,
-      "long_answer_loss": 0.0633,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 644,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0928,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.2237782786725215e-05,
-      "long_answer_loss": 0.0928,
-      "loss": 0.0798,
-      "short_answer_loss": NaN,
-      "step": 645,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.04,
-      "full_loss": 0.0697,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2205011382946024e-05,
-      "long_answer_loss": 0.0697,
-      "loss": 0.0757,
-      "short_answer_loss": NaN,
-      "step": 646,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0761,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.2172242007742066e-05,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0738,
-      "short_answer_loss": NaN,
-      "step": 647,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0855,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.21394748864615e-05,
-      "long_answer_loss": 0.0855,
-      "loss": 0.0834,
-      "short_answer_loss": NaN,
-      "step": 648,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0786,
-      "grad_norm": 1.625,
-      "learning_rate": 1.210671024443699e-05,
-      "long_answer_loss": 0.0786,
-      "loss": 0.0752,
-      "short_answer_loss": NaN,
-      "step": 649,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0761,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.2073948306984148e-05,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0794,
-      "short_answer_loss": NaN,
-      "step": 650,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.093,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.2041189299399991e-05,
-      "long_answer_loss": 0.093,
-      "loss": 0.0855,
-      "short_answer_loss": NaN,
-      "step": 651,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.05,
-      "full_loss": 0.0824,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.2008433446961384e-05,
-      "long_answer_loss": 0.0824,
-      "loss": 0.0753,
-      "short_answer_loss": NaN,
-      "step": 652,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0785,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.1975680974923497e-05,
-      "long_answer_loss": 0.0785,
-      "loss": 0.0746,
-      "short_answer_loss": NaN,
-      "step": 653,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0799,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.194293210851825e-05,
-      "long_answer_loss": 0.0799,
-      "loss": 0.0826,
-      "short_answer_loss": NaN,
-      "step": 654,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0642,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.1910187072952779e-05,
-      "long_answer_loss": 0.0642,
-      "loss": 0.0764,
-      "short_answer_loss": NaN,
-      "step": 655,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0772,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.1877446093407861e-05,
-      "long_answer_loss": 0.0772,
-      "loss": 0.0729,
-      "short_answer_loss": NaN,
-      "step": 656,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0799,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.184470939503639e-05,
-      "long_answer_loss": 0.0799,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 657,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.06,
-      "full_loss": 0.0733,
-      "grad_norm": 1.734375,
-      "learning_rate": 1.1811977202961817e-05,
-      "long_answer_loss": 0.0733,
-      "loss": 0.0795,
-      "short_answer_loss": NaN,
-      "step": 658,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.089,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.1779249742276603e-05,
-      "long_answer_loss": 0.089,
-      "loss": 0.0743,
-      "short_answer_loss": NaN,
-      "step": 659,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0853,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.1746527238040674e-05,
-      "long_answer_loss": 0.0853,
-      "loss": 0.0819,
-      "short_answer_loss": NaN,
-      "step": 660,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.069,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.171380991527987e-05,
-      "long_answer_loss": 0.069,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 661,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0942,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.1681097998984401e-05,
-      "long_answer_loss": 0.0942,
-      "loss": 0.0812,
-      "short_answer_loss": NaN,
-      "step": 662,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0842,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.1648391714107295e-05,
-      "long_answer_loss": 0.0842,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 663,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.07,
-      "full_loss": 0.0594,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.1615691285562857e-05,
-      "long_answer_loss": 0.0594,
-      "loss": 0.076,
-      "short_answer_loss": NaN,
-      "step": 664,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0882,
-      "grad_norm": 1.6640625,
-      "learning_rate": 1.1582996938225119e-05,
-      "long_answer_loss": 0.0882,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 665,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0722,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.1550308896926288e-05,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 666,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0663,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.1517627386455215e-05,
-      "long_answer_loss": 0.0663,
-      "loss": 0.0778,
-      "short_answer_loss": NaN,
-      "step": 667,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0864,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.1484952631555834e-05,
-      "long_answer_loss": 0.0864,
-      "loss": 0.0764,
-      "short_answer_loss": NaN,
-      "step": 668,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0743,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.1452284856925621e-05,
-      "long_answer_loss": 0.0743,
-      "loss": 0.0791,
-      "short_answer_loss": NaN,
-      "step": 669,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0793,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.1419624287214057e-05,
-      "long_answer_loss": 0.0793,
-      "loss": 0.0777,
-      "short_answer_loss": NaN,
-      "step": 670,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.08,
-      "full_loss": 0.0805,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.1386971147021067e-05,
-      "long_answer_loss": 0.0805,
-      "loss": 0.0801,
-      "short_answer_loss": NaN,
-      "step": 671,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0593,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.1354325660895496e-05,
-      "long_answer_loss": 0.0593,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 672,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0708,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.132168805333354e-05,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0755,
-      "short_answer_loss": NaN,
-      "step": 673,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0748,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.1289058548777229e-05,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0804,
-      "short_answer_loss": NaN,
-      "step": 674,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.089,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.125643737161286e-05,
-      "long_answer_loss": 0.089,
-      "loss": 0.0764,
-      "short_answer_loss": NaN,
-      "step": 675,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0728,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.1223824746169472e-05,
-      "long_answer_loss": 0.0728,
-      "loss": 0.0801,
-      "short_answer_loss": NaN,
-      "step": 676,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.09,
-      "full_loss": 0.0789,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.119122089671729e-05,
-      "long_answer_loss": 0.0789,
-      "loss": 0.0721,
-      "short_answer_loss": NaN,
-      "step": 677,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0764,
-      "grad_norm": 1.6484375,
-      "learning_rate": 1.1158626047466191e-05,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0802,
-      "short_answer_loss": NaN,
-      "step": 678,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.085,
-      "grad_norm": 1.5,
-      "learning_rate": 1.112604042256416e-05,
-      "long_answer_loss": 0.085,
-      "loss": 0.0775,
-      "short_answer_loss": NaN,
-      "step": 679,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0571,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.1093464246095746e-05,
-      "long_answer_loss": 0.0571,
-      "loss": 0.0771,
-      "short_answer_loss": NaN,
-      "step": 680,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0718,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.1060897742080525e-05,
-      "long_answer_loss": 0.0718,
-      "loss": 0.0747,
-      "short_answer_loss": NaN,
-      "step": 681,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0846,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.1028341134471554e-05,
-      "long_answer_loss": 0.0846,
-      "loss": 0.0819,
-      "short_answer_loss": NaN,
-      "step": 682,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.1,
-      "full_loss": 0.0797,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.0995794647153842e-05,
-      "long_answer_loss": 0.0797,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 683,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.078,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.0963258503942795e-05,
-      "long_answer_loss": 0.078,
-      "loss": 0.0824,
-      "short_answer_loss": NaN,
-      "step": 684,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.1139,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.0930732928582687e-05,
-      "long_answer_loss": 0.1139,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 685,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0878,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.0898218144745123e-05,
-      "long_answer_loss": 0.0878,
-      "loss": 0.076,
-      "short_answer_loss": NaN,
-      "step": 686,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0792,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0865714376027488e-05,
-      "long_answer_loss": 0.0792,
-      "loss": 0.0737,
-      "short_answer_loss": NaN,
-      "step": 687,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0748,
-      "grad_norm": 1.640625,
-      "learning_rate": 1.0833221845951433e-05,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 688,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.11,
-      "full_loss": 0.0648,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.080074077796131e-05,
-      "long_answer_loss": 0.0648,
-      "loss": 0.0709,
-      "short_answer_loss": NaN,
-      "step": 689,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0664,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.0768271395422651e-05,
-      "long_answer_loss": 0.0664,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 690,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0841,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.0735813921620634e-05,
-      "long_answer_loss": 0.0841,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 691,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.079,
-      "grad_norm": 1.6796875,
-      "learning_rate": 1.070336857975854e-05,
-      "long_answer_loss": 0.079,
-      "loss": 0.0814,
-      "short_answer_loss": NaN,
-      "step": 692,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0902,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.0670935592956223e-05,
-      "long_answer_loss": 0.0902,
-      "loss": 0.0773,
-      "short_answer_loss": NaN,
-      "step": 693,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0867,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.0638515184248571e-05,
-      "long_answer_loss": 0.0867,
-      "loss": 0.0734,
-      "short_answer_loss": NaN,
-      "step": 694,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.12,
-      "full_loss": 0.0694,
-      "grad_norm": 1.5,
-      "learning_rate": 1.0606107576583976e-05,
-      "long_answer_loss": 0.0694,
-      "loss": 0.0738,
-      "short_answer_loss": NaN,
-      "step": 695,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.1009,
-      "grad_norm": 1.59375,
-      "learning_rate": 1.0573712992822804e-05,
-      "long_answer_loss": 0.1009,
-      "loss": 0.0839,
-      "short_answer_loss": NaN,
-      "step": 696,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0949,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.0541331655735853e-05,
-      "long_answer_loss": 0.0949,
-      "loss": 0.0754,
-      "short_answer_loss": NaN,
-      "step": 697,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0804,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.0508963788002827e-05,
-      "long_answer_loss": 0.0804,
-      "loss": 0.0738,
-      "short_answer_loss": NaN,
-      "step": 698,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0741,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.0476609612210808e-05,
-      "long_answer_loss": 0.0741,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 699,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0814,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.0444269350852718e-05,
-      "long_answer_loss": 0.0814,
-      "loss": 0.0791,
-      "short_answer_loss": NaN,
-      "step": 700,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.13,
-      "full_loss": 0.0724,
-      "grad_norm": 1.578125,
-      "learning_rate": 1.0411943226325793e-05,
-      "long_answer_loss": 0.0724,
-      "loss": 0.071,
-      "short_answer_loss": NaN,
-      "step": 701,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.084,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.0379631460930054e-05,
-      "long_answer_loss": 0.084,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 702,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0656,
-      "grad_norm": 1.671875,
-      "learning_rate": 1.0347334276866772e-05,
-      "long_answer_loss": 0.0656,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 703,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0907,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.0315051896236955e-05,
-      "long_answer_loss": 0.0907,
-      "loss": 0.0867,
-      "short_answer_loss": NaN,
-      "step": 704,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.073,
-      "grad_norm": 1.625,
-      "learning_rate": 1.0282784541039804e-05,
-      "long_answer_loss": 0.073,
-      "loss": 0.0808,
-      "short_answer_loss": NaN,
-      "step": 705,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0825,
-      "grad_norm": 1.6328125,
-      "learning_rate": 1.0250532433171194e-05,
-      "long_answer_loss": 0.0825,
-      "loss": 0.0803,
-      "short_answer_loss": NaN,
-      "step": 706,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.0669,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.0218295794422147e-05,
-      "long_answer_loss": 0.0669,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 707,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.14,
-      "full_loss": 0.08,
-      "grad_norm": 1.65625,
-      "learning_rate": 1.018607484647731e-05,
-      "long_answer_loss": 0.08,
-      "loss": 0.0771,
-      "short_answer_loss": NaN,
-      "step": 708,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0723,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.0153869810913424e-05,
-      "long_answer_loss": 0.0723,
-      "loss": 0.0799,
-      "short_answer_loss": NaN,
-      "step": 709,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0776,
-      "grad_norm": 1.6171875,
-      "learning_rate": 1.0121680909197809e-05,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0761,
-      "short_answer_loss": NaN,
-      "step": 710,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0681,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.0089508362686827e-05,
-      "long_answer_loss": 0.0681,
-      "loss": 0.0792,
-      "short_answer_loss": NaN,
-      "step": 711,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0799,
-      "grad_norm": 1.546875,
-      "learning_rate": 1.0057352392624377e-05,
-      "long_answer_loss": 0.0799,
-      "loss": 0.079,
-      "short_answer_loss": NaN,
-      "step": 712,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.074,
-      "grad_norm": 1.6015625,
-      "learning_rate": 1.0025213220140364e-05,
-      "long_answer_loss": 0.074,
-      "loss": 0.0752,
-      "short_answer_loss": NaN,
-      "step": 713,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.15,
-      "full_loss": 0.0629,
-      "grad_norm": 1.578125,
-      "learning_rate": 9.993091066249174e-06,
-      "long_answer_loss": 0.0629,
-      "loss": 0.0749,
-      "short_answer_loss": NaN,
-      "step": 714,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0891,
-      "grad_norm": 1.5546875,
-      "learning_rate": 9.960986151848167e-06,
-      "long_answer_loss": 0.0891,
-      "loss": 0.0742,
-      "short_answer_loss": NaN,
-      "step": 715,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0686,
-      "grad_norm": 1.5,
-      "learning_rate": 9.928898697716147e-06,
-      "long_answer_loss": 0.0686,
-      "loss": 0.0675,
-      "short_answer_loss": NaN,
-      "step": 716,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0818,
-      "grad_norm": 1.453125,
-      "learning_rate": 9.896828924511845e-06,
-      "long_answer_loss": 0.0818,
-      "loss": 0.0782,
-      "short_answer_loss": NaN,
-      "step": 717,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0696,
-      "grad_norm": 1.53125,
-      "learning_rate": 9.864777052772407e-06,
-      "long_answer_loss": 0.0696,
-      "loss": 0.077,
-      "short_answer_loss": NaN,
-      "step": 718,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0777,
-      "grad_norm": 1.5546875,
-      "learning_rate": 9.832743302911876e-06,
-      "long_answer_loss": 0.0777,
-      "loss": 0.0785,
-      "short_answer_loss": NaN,
-      "step": 719,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.16,
-      "full_loss": 0.0767,
-      "grad_norm": 1.5703125,
-      "learning_rate": 9.800727895219672e-06,
-      "long_answer_loss": 0.0767,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 720,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0671,
-      "grad_norm": 1.5,
-      "learning_rate": 9.768731049859073e-06,
-      "long_answer_loss": 0.0671,
-      "loss": 0.0729,
-      "short_answer_loss": NaN,
-      "step": 721,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.072,
-      "grad_norm": 1.546875,
-      "learning_rate": 9.736752986865727e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0773,
-      "short_answer_loss": NaN,
-      "step": 722,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0713,
-      "grad_norm": 1.5078125,
-      "learning_rate": 9.704793926146102e-06,
-      "long_answer_loss": 0.0713,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 723,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0717,
-      "grad_norm": 1.484375,
-      "learning_rate": 9.672854087475997e-06,
-      "long_answer_loss": 0.0717,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 724,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.072,
-      "grad_norm": 1.5078125,
-      "learning_rate": 9.640933690499027e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 725,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.17,
-      "full_loss": 0.0798,
-      "grad_norm": 1.53125,
-      "learning_rate": 9.609032954725104e-06,
-      "long_answer_loss": 0.0798,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 726,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0731,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.57715209952894e-06,
-      "long_answer_loss": 0.0731,
-      "loss": 0.069,
-      "short_answer_loss": NaN,
-      "step": 727,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0856,
-      "grad_norm": 1.4453125,
-      "learning_rate": 9.54529134414853e-06,
-      "long_answer_loss": 0.0856,
-      "loss": 0.0762,
-      "short_answer_loss": NaN,
-      "step": 728,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0758,
-      "grad_norm": 1.6875,
-      "learning_rate": 9.51345090768365e-06,
-      "long_answer_loss": 0.0758,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 729,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0776,
-      "grad_norm": 1.5234375,
-      "learning_rate": 9.481631009094341e-06,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0745,
-      "short_answer_loss": NaN,
-      "step": 730,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0583,
-      "grad_norm": 1.5078125,
-      "learning_rate": 9.449831867199416e-06,
-      "long_answer_loss": 0.0583,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 731,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.18,
-      "full_loss": 0.0661,
-      "grad_norm": 1.4921875,
-      "learning_rate": 9.418053700674944e-06,
-      "long_answer_loss": 0.0661,
-      "loss": 0.0754,
-      "short_answer_loss": NaN,
-      "step": 732,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0722,
-      "grad_norm": 1.6015625,
-      "learning_rate": 9.386296728052753e-06,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 733,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0594,
-      "grad_norm": 1.6328125,
-      "learning_rate": 9.354561167718922e-06,
-      "long_answer_loss": 0.0594,
-      "loss": 0.0725,
-      "short_answer_loss": NaN,
-      "step": 734,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0713,
-      "grad_norm": 1.65625,
-      "learning_rate": 9.322847237912288e-06,
-      "long_answer_loss": 0.0713,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 735,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.1066,
-      "grad_norm": 1.5859375,
-      "learning_rate": 9.29115515672293e-06,
-      "long_answer_loss": 0.1066,
-      "loss": 0.0806,
-      "short_answer_loss": NaN,
-      "step": 736,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0749,
-      "grad_norm": 1.609375,
-      "learning_rate": 9.25948514209069e-06,
-      "long_answer_loss": 0.0749,
-      "loss": 0.0761,
-      "short_answer_loss": NaN,
-      "step": 737,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0544,
-      "grad_norm": 1.5703125,
-      "learning_rate": 9.227837411803656e-06,
-      "long_answer_loss": 0.0544,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 738,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.19,
-      "full_loss": 0.0895,
-      "grad_norm": 1.546875,
-      "learning_rate": 9.196212183496669e-06,
-      "long_answer_loss": 0.0895,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 739,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.1009,
-      "grad_norm": 1.546875,
-      "learning_rate": 9.164609674649835e-06,
-      "long_answer_loss": 0.1009,
-      "loss": 0.0761,
-      "short_answer_loss": NaN,
-      "step": 740,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0684,
-      "grad_norm": 1.6171875,
-      "learning_rate": 9.133030102587019e-06,
-      "long_answer_loss": 0.0684,
-      "loss": 0.0754,
-      "short_answer_loss": NaN,
-      "step": 741,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0647,
-      "grad_norm": 1.5703125,
-      "learning_rate": 9.101473684474354e-06,
-      "long_answer_loss": 0.0647,
-      "loss": 0.0785,
-      "short_answer_loss": NaN,
-      "step": 742,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0748,
-      "grad_norm": 1.4609375,
-      "learning_rate": 9.069940637318752e-06,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 743,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0662,
-      "grad_norm": 1.46875,
-      "learning_rate": 9.038431177966406e-06,
-      "long_answer_loss": 0.0662,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 744,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.2,
-      "full_loss": 0.0857,
-      "grad_norm": 1.4453125,
-      "learning_rate": 9.006945523101295e-06,
-      "long_answer_loss": 0.0857,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 745,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0737,
-      "grad_norm": 1.484375,
-      "learning_rate": 8.975483889243709e-06,
-      "long_answer_loss": 0.0737,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 746,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0732,
-      "grad_norm": 1.4609375,
-      "learning_rate": 8.944046492748746e-06,
-      "long_answer_loss": 0.0732,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 747,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0683,
-      "grad_norm": 1.546875,
-      "learning_rate": 8.912633549804824e-06,
-      "long_answer_loss": 0.0683,
-      "loss": 0.0704,
-      "short_answer_loss": NaN,
-      "step": 748,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0797,
-      "grad_norm": 1.484375,
-      "learning_rate": 8.88124527643221e-06,
-      "long_answer_loss": 0.0797,
-      "loss": 0.0754,
-      "short_answer_loss": NaN,
-      "step": 749,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0627,
-      "grad_norm": 1.5,
-      "learning_rate": 8.849881888481513e-06,
-      "long_answer_loss": 0.0627,
-      "loss": 0.0738,
-      "short_answer_loss": NaN,
-      "step": 750,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.21,
-      "full_loss": 0.0746,
-      "grad_norm": 1.5546875,
-      "learning_rate": 8.818543601632215e-06,
-      "long_answer_loss": 0.0746,
-      "loss": 0.0736,
-      "short_answer_loss": NaN,
-      "step": 751,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0773,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.787230631391185e-06,
-      "long_answer_loss": 0.0773,
-      "loss": 0.0713,
-      "short_answer_loss": NaN,
-      "step": 752,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0852,
-      "grad_norm": 1.5703125,
-      "learning_rate": 8.755943193091187e-06,
-      "long_answer_loss": 0.0852,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 753,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0669,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.724681501889413e-06,
-      "long_answer_loss": 0.0669,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 754,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0729,
-      "grad_norm": 1.609375,
-      "learning_rate": 8.693445772766003e-06,
-      "long_answer_loss": 0.0729,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 755,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0827,
-      "grad_norm": 1.453125,
-      "learning_rate": 8.662236220522554e-06,
-      "long_answer_loss": 0.0827,
-      "loss": 0.0755,
-      "short_answer_loss": NaN,
-      "step": 756,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.22,
-      "full_loss": 0.0792,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.631053059780647e-06,
-      "long_answer_loss": 0.0792,
-      "loss": 0.0745,
-      "short_answer_loss": NaN,
-      "step": 757,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0687,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.599896504980384e-06,
-      "long_answer_loss": 0.0687,
-      "loss": 0.0755,
-      "short_answer_loss": NaN,
-      "step": 758,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0661,
-      "grad_norm": 1.4296875,
-      "learning_rate": 8.568766770378892e-06,
-      "long_answer_loss": 0.0661,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 759,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0701,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.537664070048867e-06,
-      "long_answer_loss": 0.0701,
-      "loss": 0.0723,
-      "short_answer_loss": NaN,
-      "step": 760,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0782,
-      "grad_norm": 1.5078125,
-      "learning_rate": 8.506588617877102e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0786,
-      "short_answer_loss": NaN,
-      "step": 761,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0597,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.475540627563e-06,
-      "long_answer_loss": 0.0597,
-      "loss": 0.0724,
-      "short_answer_loss": NaN,
-      "step": 762,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.23,
-      "full_loss": 0.0623,
-      "grad_norm": 1.5,
-      "learning_rate": 8.444520312617118e-06,
-      "long_answer_loss": 0.0623,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 763,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0693,
-      "grad_norm": 1.4921875,
-      "learning_rate": 8.413527886359695e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.074,
-      "short_answer_loss": NaN,
-      "step": 764,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0785,
-      "grad_norm": 1.4296875,
-      "learning_rate": 8.382563561919191e-06,
-      "long_answer_loss": 0.0785,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 765,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0677,
-      "grad_norm": 1.6171875,
-      "learning_rate": 8.351627552230806e-06,
-      "long_answer_loss": 0.0677,
-      "loss": 0.0747,
-      "short_answer_loss": NaN,
-      "step": 766,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0751,
-      "grad_norm": 1.609375,
-      "learning_rate": 8.320720070035035e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0776,
-      "short_answer_loss": NaN,
-      "step": 767,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0693,
-      "grad_norm": 1.53125,
-      "learning_rate": 8.289841327876183e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0737,
-      "short_answer_loss": NaN,
-      "step": 768,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0721,
-      "grad_norm": 1.5234375,
-      "learning_rate": 8.25899153810093e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0712,
-      "short_answer_loss": NaN,
-      "step": 769,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.24,
-      "full_loss": 0.0587,
-      "grad_norm": 1.515625,
-      "learning_rate": 8.228170912856845e-06,
-      "long_answer_loss": 0.0587,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 770,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0852,
-      "grad_norm": 1.5078125,
-      "learning_rate": 8.197379664090947e-06,
-      "long_answer_loss": 0.0852,
-      "loss": 0.072,
-      "short_answer_loss": NaN,
-      "step": 771,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0758,
-      "grad_norm": 1.6015625,
-      "learning_rate": 8.166618003548235e-06,
-      "long_answer_loss": 0.0758,
-      "loss": 0.0774,
-      "short_answer_loss": NaN,
-      "step": 772,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0672,
-      "grad_norm": 1.40625,
-      "learning_rate": 8.135886142770232e-06,
-      "long_answer_loss": 0.0672,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 773,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0588,
-      "grad_norm": 1.453125,
-      "learning_rate": 8.105184293093545e-06,
-      "long_answer_loss": 0.0588,
-      "loss": 0.0678,
-      "short_answer_loss": NaN,
-      "step": 774,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.0543,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.074512665648392e-06,
-      "long_answer_loss": 0.0543,
-      "loss": 0.074,
-      "short_answer_loss": NaN,
-      "step": 775,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.25,
-      "full_loss": 0.076,
-      "grad_norm": 1.515625,
-      "learning_rate": 8.04387147135716e-06,
-      "long_answer_loss": 0.076,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 776,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0697,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.013260920932957e-06,
-      "long_answer_loss": 0.0697,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 777,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.085,
-      "grad_norm": 1.46875,
-      "learning_rate": 7.982681224878157e-06,
-      "long_answer_loss": 0.085,
-      "loss": 0.0701,
-      "short_answer_loss": NaN,
-      "step": 778,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0754,
-      "grad_norm": 1.5546875,
-      "learning_rate": 7.952132593482956e-06,
-      "long_answer_loss": 0.0754,
-      "loss": 0.0753,
-      "short_answer_loss": NaN,
-      "step": 779,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0818,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.921615236823924e-06,
-      "long_answer_loss": 0.0818,
-      "loss": 0.0676,
-      "short_answer_loss": NaN,
-      "step": 780,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0902,
-      "grad_norm": 1.53125,
-      "learning_rate": 7.891129364762559e-06,
-      "long_answer_loss": 0.0902,
-      "loss": 0.0805,
-      "short_answer_loss": NaN,
-      "step": 781,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.26,
-      "full_loss": 0.0507,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.860675186943853e-06,
-      "long_answer_loss": 0.0507,
-      "loss": 0.0654,
-      "short_answer_loss": NaN,
-      "step": 782,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0574,
-      "grad_norm": 1.5,
-      "learning_rate": 7.830252912794836e-06,
-      "long_answer_loss": 0.0574,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 783,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0782,
-      "grad_norm": 1.59375,
-      "learning_rate": 7.799862751523146e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 784,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0597,
-      "grad_norm": 1.4765625,
-      "learning_rate": 7.769504912115588e-06,
-      "long_answer_loss": 0.0597,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 785,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0651,
-      "grad_norm": 1.46875,
-      "learning_rate": 7.739179603336696e-06,
-      "long_answer_loss": 0.0651,
-      "loss": 0.0726,
-      "short_answer_loss": NaN,
-      "step": 786,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0657,
-      "grad_norm": 1.4296875,
-      "learning_rate": 7.708887033727291e-06,
-      "long_answer_loss": 0.0657,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 787,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.27,
-      "full_loss": 0.0599,
-      "grad_norm": 1.40625,
-      "learning_rate": 7.678627411603074e-06,
-      "long_answer_loss": 0.0599,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 788,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0741,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.648400945053146e-06,
-      "long_answer_loss": 0.0741,
-      "loss": 0.0701,
-      "short_answer_loss": NaN,
-      "step": 789,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0807,
-      "grad_norm": 1.5546875,
-      "learning_rate": 7.618207841938624e-06,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 790,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0706,
-      "grad_norm": 1.53125,
-      "learning_rate": 7.588048309891181e-06,
-      "long_answer_loss": 0.0706,
-      "loss": 0.0712,
-      "short_answer_loss": NaN,
-      "step": 791,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0788,
-      "grad_norm": 1.4609375,
-      "learning_rate": 7.557922556311634e-06,
-      "long_answer_loss": 0.0788,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 792,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0771,
-      "grad_norm": 1.4609375,
-      "learning_rate": 7.527830788368509e-06,
-      "long_answer_loss": 0.0771,
-      "loss": 0.0732,
-      "short_answer_loss": NaN,
-      "step": 793,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.28,
-      "full_loss": 0.0605,
-      "grad_norm": 1.6484375,
-      "learning_rate": 7.497773212996623e-06,
-      "long_answer_loss": 0.0605,
-      "loss": 0.0793,
-      "short_answer_loss": NaN,
-      "step": 794,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0751,
-      "grad_norm": 1.59375,
-      "learning_rate": 7.467750036895657e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0742,
-      "short_answer_loss": NaN,
-      "step": 795,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0719,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.437761466528731e-06,
-      "long_answer_loss": 0.0719,
-      "loss": 0.0717,
-      "short_answer_loss": NaN,
-      "step": 796,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0624,
-      "grad_norm": 1.5234375,
-      "learning_rate": 7.407807708120998e-06,
-      "long_answer_loss": 0.0624,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 797,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0642,
-      "grad_norm": 1.4140625,
-      "learning_rate": 7.377888967658206e-06,
-      "long_answer_loss": 0.0642,
-      "loss": 0.0701,
-      "short_answer_loss": NaN,
-      "step": 798,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0696,
-      "grad_norm": 1.5390625,
-      "learning_rate": 7.348005450885301e-06,
-      "long_answer_loss": 0.0696,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 799,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.29,
-      "full_loss": 0.0766,
-      "grad_norm": 1.5234375,
-      "learning_rate": 7.318157363304995e-06,
-      "long_answer_loss": 0.0766,
-      "loss": 0.071,
-      "short_answer_loss": NaN,
-      "step": 800,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0667,
-      "grad_norm": 1.5390625,
-      "learning_rate": 7.288344910176365e-06,
-      "long_answer_loss": 0.0667,
-      "loss": 0.0734,
-      "short_answer_loss": NaN,
-      "step": 801,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0701,
-      "grad_norm": 1.5625,
-      "learning_rate": 7.258568296513439e-06,
-      "long_answer_loss": 0.0701,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 802,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0903,
-      "grad_norm": 1.6171875,
-      "learning_rate": 7.228827727083781e-06,
-      "long_answer_loss": 0.0903,
-      "loss": 0.0759,
-      "short_answer_loss": NaN,
-      "step": 803,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0726,
-      "grad_norm": 1.53125,
-      "learning_rate": 7.199123406407089e-06,
-      "long_answer_loss": 0.0726,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 804,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0762,
-      "grad_norm": 1.5,
-      "learning_rate": 7.169455538753783e-06,
-      "long_answer_loss": 0.0762,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 805,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0821,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.139824328143604e-06,
-      "long_answer_loss": 0.0821,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 806,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.3,
-      "full_loss": 0.0701,
-      "grad_norm": 1.4765625,
-      "learning_rate": 7.110229978344212e-06,
-      "long_answer_loss": 0.0701,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 807,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0829,
-      "grad_norm": 1.5625,
-      "learning_rate": 7.080672692869783e-06,
-      "long_answer_loss": 0.0829,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 808,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0703,
-      "grad_norm": 1.40625,
-      "learning_rate": 7.051152674979608e-06,
-      "long_answer_loss": 0.0703,
-      "loss": 0.0672,
-      "short_answer_loss": NaN,
-      "step": 809,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0737,
-      "grad_norm": 1.5078125,
-      "learning_rate": 7.0216701276766936e-06,
-      "long_answer_loss": 0.0737,
-      "loss": 0.0744,
-      "short_answer_loss": NaN,
-      "step": 810,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0751,
-      "grad_norm": 1.5546875,
-      "learning_rate": 6.992225253706374e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 811,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0662,
-      "grad_norm": 1.625,
-      "learning_rate": 6.962818255554911e-06,
-      "long_answer_loss": 0.0662,
-      "loss": 0.0747,
-      "short_answer_loss": NaN,
-      "step": 812,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.31,
-      "full_loss": 0.0629,
-      "grad_norm": 1.4765625,
-      "learning_rate": 6.9334493354480985e-06,
-      "long_answer_loss": 0.0629,
-      "loss": 0.0684,
-      "short_answer_loss": NaN,
-      "step": 813,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0756,
-      "grad_norm": 1.5234375,
-      "learning_rate": 6.904118695349882e-06,
-      "long_answer_loss": 0.0756,
-      "loss": 0.0747,
-      "short_answer_loss": NaN,
-      "step": 814,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0698,
-      "grad_norm": 1.5390625,
-      "learning_rate": 6.874826536960954e-06,
-      "long_answer_loss": 0.0698,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 815,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0696,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.845573061717387e-06,
-      "long_answer_loss": 0.0696,
-      "loss": 0.0777,
-      "short_answer_loss": NaN,
-      "step": 816,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0788,
-      "grad_norm": 1.5546875,
-      "learning_rate": 6.8163584707892306e-06,
-      "long_answer_loss": 0.0788,
-      "loss": 0.0712,
-      "short_answer_loss": NaN,
-      "step": 817,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0925,
-      "grad_norm": 1.6171875,
-      "learning_rate": 6.7871829650791365e-06,
-      "long_answer_loss": 0.0925,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 818,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.32,
-      "full_loss": 0.0659,
-      "grad_norm": 1.5703125,
-      "learning_rate": 6.758046745220978e-06,
-      "long_answer_loss": 0.0659,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 819,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.063,
-      "grad_norm": 1.59375,
-      "learning_rate": 6.728950011578462e-06,
-      "long_answer_loss": 0.063,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 820,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0639,
-      "grad_norm": 1.578125,
-      "learning_rate": 6.6998929642437645e-06,
-      "long_answer_loss": 0.0639,
-      "loss": 0.0731,
-      "short_answer_loss": NaN,
-      "step": 821,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0546,
-      "grad_norm": 1.421875,
-      "learning_rate": 6.670875803036141e-06,
-      "long_answer_loss": 0.0546,
-      "loss": 0.07,
-      "short_answer_loss": NaN,
-      "step": 822,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0727,
-      "grad_norm": 1.5625,
-      "learning_rate": 6.64189872750056e-06,
-      "long_answer_loss": 0.0727,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 823,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0826,
-      "grad_norm": 1.4765625,
-      "learning_rate": 6.612961936906333e-06,
-      "long_answer_loss": 0.0826,
-      "loss": 0.074,
-      "short_answer_loss": NaN,
-      "step": 824,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.33,
-      "full_loss": 0.0692,
-      "grad_norm": 1.4453125,
-      "learning_rate": 6.584065630245734e-06,
-      "long_answer_loss": 0.0692,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 825,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0577,
-      "grad_norm": 1.4921875,
-      "learning_rate": 6.55521000623264e-06,
-      "long_answer_loss": 0.0577,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 826,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0838,
-      "grad_norm": 1.515625,
-      "learning_rate": 6.526395263301166e-06,
-      "long_answer_loss": 0.0838,
-      "loss": 0.0716,
-      "short_answer_loss": NaN,
-      "step": 827,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0701,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.497621599604292e-06,
-      "long_answer_loss": 0.0701,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 828,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.082,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.468889213012502e-06,
-      "long_answer_loss": 0.082,
-      "loss": 0.0712,
-      "short_answer_loss": NaN,
-      "step": 829,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0891,
-      "grad_norm": 1.4375,
-      "learning_rate": 6.440198301112434e-06,
-      "long_answer_loss": 0.0891,
-      "loss": 0.0723,
-      "short_answer_loss": NaN,
-      "step": 830,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.34,
-      "full_loss": 0.0972,
-      "grad_norm": 1.5859375,
-      "learning_rate": 6.411549061205505e-06,
-      "long_answer_loss": 0.0972,
-      "loss": 0.0753,
-      "short_answer_loss": NaN,
-      "step": 831,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0513,
-      "grad_norm": 1.5703125,
-      "learning_rate": 6.382941690306568e-06,
-      "long_answer_loss": 0.0513,
-      "loss": 0.0653,
-      "short_answer_loss": NaN,
-      "step": 832,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0743,
-      "grad_norm": 1.4765625,
-      "learning_rate": 6.35437638514255e-06,
-      "long_answer_loss": 0.0743,
-      "loss": 0.0673,
-      "short_answer_loss": NaN,
-      "step": 833,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0729,
-      "grad_norm": 1.390625,
-      "learning_rate": 6.325853342151097e-06,
-      "long_answer_loss": 0.0729,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 834,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0751,
-      "grad_norm": 1.6484375,
-      "learning_rate": 6.2973727574792345e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0704,
-      "short_answer_loss": NaN,
-      "step": 835,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0665,
-      "grad_norm": 1.6640625,
-      "learning_rate": 6.2689348269820036e-06,
-      "long_answer_loss": 0.0665,
-      "loss": 0.0742,
-      "short_answer_loss": NaN,
-      "step": 836,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0658,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.240539746221127e-06,
-      "long_answer_loss": 0.0658,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 837,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.35,
-      "full_loss": 0.0639,
-      "grad_norm": 1.4375,
-      "learning_rate": 6.212187710463654e-06,
-      "long_answer_loss": 0.0639,
-      "loss": 0.075,
-      "short_answer_loss": NaN,
-      "step": 838,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0708,
-      "grad_norm": 1.5703125,
-      "learning_rate": 6.1838789146806254e-06,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0681,
-      "short_answer_loss": NaN,
-      "step": 839,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0712,
-      "grad_norm": 1.5859375,
-      "learning_rate": 6.155613553545729e-06,
-      "long_answer_loss": 0.0712,
-      "loss": 0.0749,
-      "short_answer_loss": NaN,
-      "step": 840,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0751,
-      "grad_norm": 1.40625,
-      "learning_rate": 6.127391821433961e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0717,
-      "short_answer_loss": NaN,
-      "step": 841,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0764,
-      "grad_norm": 1.5546875,
-      "learning_rate": 6.0992139124202914e-06,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 842,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0701,
-      "grad_norm": 1.5703125,
-      "learning_rate": 6.071080020278326e-06,
-      "long_answer_loss": 0.0701,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 843,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.36,
-      "full_loss": 0.0805,
-      "grad_norm": 1.609375,
-      "learning_rate": 6.0429903384789775e-06,
-      "long_answer_loss": 0.0805,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 844,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0748,
-      "grad_norm": 1.6640625,
-      "learning_rate": 6.0149450601891325e-06,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0817,
-      "short_answer_loss": NaN,
-      "step": 845,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0626,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.986944378270323e-06,
-      "long_answer_loss": 0.0626,
-      "loss": 0.0746,
-      "short_answer_loss": NaN,
-      "step": 846,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0844,
-      "grad_norm": 1.578125,
-      "learning_rate": 5.958988485277401e-06,
-      "long_answer_loss": 0.0844,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 847,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0873,
-      "grad_norm": 1.71875,
-      "learning_rate": 5.93107757345722e-06,
-      "long_answer_loss": 0.0873,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 848,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0692,
-      "grad_norm": 1.453125,
-      "learning_rate": 5.9032118347472965e-06,
-      "long_answer_loss": 0.0692,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 849,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.37,
-      "full_loss": 0.0654,
-      "grad_norm": 1.453125,
-      "learning_rate": 5.87539146077451e-06,
-      "long_answer_loss": 0.0654,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 850,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0774,
-      "grad_norm": 1.4765625,
-      "learning_rate": 5.847616642853773e-06,
-      "long_answer_loss": 0.0774,
-      "loss": 0.069,
-      "short_answer_loss": NaN,
-      "step": 851,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0641,
-      "grad_norm": 1.5234375,
-      "learning_rate": 5.81988757198672e-06,
-      "long_answer_loss": 0.0641,
-      "loss": 0.0725,
-      "short_answer_loss": NaN,
-      "step": 852,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0771,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.792204438860391e-06,
-      "long_answer_loss": 0.0771,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 853,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0661,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.764567433845915e-06,
-      "long_answer_loss": 0.0661,
-      "loss": 0.07,
-      "short_answer_loss": NaN,
-      "step": 854,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0721,
-      "grad_norm": 1.5,
-      "learning_rate": 5.736976746997226e-06,
-      "long_answer_loss": 0.0721,
-      "loss": 0.069,
-      "short_answer_loss": NaN,
-      "step": 855,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.38,
-      "full_loss": 0.0943,
-      "grad_norm": 1.453125,
-      "learning_rate": 5.709432568049722e-06,
-      "long_answer_loss": 0.0943,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 856,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0672,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.681935086418978e-06,
-      "long_answer_loss": 0.0672,
-      "loss": 0.0715,
-      "short_answer_loss": NaN,
-      "step": 857,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0625,
-      "grad_norm": 1.6328125,
-      "learning_rate": 5.654484491199446e-06,
-      "long_answer_loss": 0.0625,
-      "loss": 0.0724,
-      "short_answer_loss": NaN,
-      "step": 858,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0597,
-      "grad_norm": 1.5,
-      "learning_rate": 5.627080971163146e-06,
-      "long_answer_loss": 0.0597,
-      "loss": 0.0674,
-      "short_answer_loss": NaN,
-      "step": 859,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0597,
-      "grad_norm": 1.4921875,
-      "learning_rate": 5.599724714758374e-06,
-      "long_answer_loss": 0.0597,
-      "loss": 0.0716,
-      "short_answer_loss": NaN,
-      "step": 860,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0699,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.572415910108401e-06,
-      "long_answer_loss": 0.0699,
-      "loss": 0.0742,
-      "short_answer_loss": NaN,
-      "step": 861,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.39,
-      "full_loss": 0.0782,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.545154745010187e-06,
-      "long_answer_loss": 0.0782,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 862,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0752,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.5179414069330786e-06,
-      "long_answer_loss": 0.0752,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 863,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.06,
-      "grad_norm": 1.375,
-      "learning_rate": 5.490776083017532e-06,
-      "long_answer_loss": 0.06,
-      "loss": 0.0691,
-      "short_answer_loss": NaN,
-      "step": 864,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0703,
-      "grad_norm": 1.5078125,
-      "learning_rate": 5.463658960073816e-06,
-      "long_answer_loss": 0.0703,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 865,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.063,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.436590224580733e-06,
-      "long_answer_loss": 0.063,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 866,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0656,
-      "grad_norm": 1.4296875,
-      "learning_rate": 5.409570062684334e-06,
-      "long_answer_loss": 0.0656,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 867,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.4,
-      "full_loss": 0.0677,
-      "grad_norm": 1.4921875,
-      "learning_rate": 5.382598660196642e-06,
-      "long_answer_loss": 0.0677,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 868,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0569,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.355676202594367e-06,
-      "long_answer_loss": 0.0569,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 869,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0708,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.3288028750176395e-06,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 870,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0767,
-      "grad_norm": 1.4609375,
-      "learning_rate": 5.301978862268733e-06,
-      "long_answer_loss": 0.0767,
-      "loss": 0.0732,
-      "short_answer_loss": NaN,
-      "step": 871,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0593,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.275204348810789e-06,
-      "long_answer_loss": 0.0593,
-      "loss": 0.0681,
-      "short_answer_loss": NaN,
-      "step": 872,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0584,
-      "grad_norm": 1.3984375,
-      "learning_rate": 5.248479518766558e-06,
-      "long_answer_loss": 0.0584,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 873,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0605,
-      "grad_norm": 1.390625,
-      "learning_rate": 5.221804555917123e-06,
-      "long_answer_loss": 0.0605,
-      "loss": 0.0648,
-      "short_answer_loss": NaN,
-      "step": 874,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.41,
-      "full_loss": 0.0637,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.195179643700646e-06,
-      "long_answer_loss": 0.0637,
-      "loss": 0.0661,
-      "short_answer_loss": NaN,
-      "step": 875,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0591,
-      "grad_norm": 1.5390625,
-      "learning_rate": 5.168604965211096e-06,
-      "long_answer_loss": 0.0591,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 876,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0751,
-      "grad_norm": 1.5234375,
-      "learning_rate": 5.142080703197e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.0725,
-      "short_answer_loss": NaN,
-      "step": 877,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0624,
-      "grad_norm": 1.4765625,
-      "learning_rate": 5.115607040060177e-06,
-      "long_answer_loss": 0.0624,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 878,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.0758,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.089184157854491e-06,
-      "long_answer_loss": 0.0758,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 879,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.06,
-      "grad_norm": 1.578125,
-      "learning_rate": 5.0628122382845935e-06,
-      "long_answer_loss": 0.06,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 880,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.42,
-      "full_loss": 0.069,
-      "grad_norm": 1.4765625,
-      "learning_rate": 5.036491462704682e-06,
-      "long_answer_loss": 0.069,
-      "loss": 0.077,
-      "short_answer_loss": NaN,
-      "step": 881,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0749,
-      "grad_norm": 1.5625,
-      "learning_rate": 5.010222012117238e-06,
-      "long_answer_loss": 0.0749,
-      "loss": 0.0736,
-      "short_answer_loss": NaN,
-      "step": 882,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0579,
-      "grad_norm": 1.5234375,
-      "learning_rate": 4.984004067171803e-06,
-      "long_answer_loss": 0.0579,
-      "loss": 0.0701,
-      "short_answer_loss": NaN,
-      "step": 883,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0791,
-      "grad_norm": 1.4140625,
-      "learning_rate": 4.957837808163718e-06,
-      "long_answer_loss": 0.0791,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 884,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0586,
-      "grad_norm": 1.546875,
-      "learning_rate": 4.931723415032889e-06,
-      "long_answer_loss": 0.0586,
-      "loss": 0.0723,
-      "short_answer_loss": NaN,
-      "step": 885,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.0948,
-      "grad_norm": 1.46875,
-      "learning_rate": 4.905661067362558e-06,
-      "long_answer_loss": 0.0948,
-      "loss": 0.0772,
-      "short_answer_loss": NaN,
-      "step": 886,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.43,
-      "full_loss": 0.072,
-      "grad_norm": 1.4140625,
-      "learning_rate": 4.87965094437805e-06,
-      "long_answer_loss": 0.072,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 887,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0702,
-      "grad_norm": 1.53125,
-      "learning_rate": 4.853693224945569e-06,
-      "long_answer_loss": 0.0702,
-      "loss": 0.0729,
-      "short_answer_loss": NaN,
-      "step": 888,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0709,
-      "grad_norm": 1.3671875,
-      "learning_rate": 4.827788087570936e-06,
-      "long_answer_loss": 0.0709,
-      "loss": 0.0689,
-      "short_answer_loss": NaN,
-      "step": 889,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0619,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.801935710398382e-06,
-      "long_answer_loss": 0.0619,
-      "loss": 0.0666,
-      "short_answer_loss": NaN,
-      "step": 890,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0726,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.776136271209315e-06,
-      "long_answer_loss": 0.0726,
-      "loss": 0.0689,
-      "short_answer_loss": NaN,
-      "step": 891,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.074,
-      "grad_norm": 1.5390625,
-      "learning_rate": 4.750389947421101e-06,
-      "long_answer_loss": 0.074,
-      "loss": 0.075,
-      "short_answer_loss": NaN,
-      "step": 892,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.44,
-      "full_loss": 0.0644,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.724696916085841e-06,
-      "long_answer_loss": 0.0644,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 893,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0673,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.699057353889157e-06,
-      "long_answer_loss": 0.0673,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 894,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0895,
-      "grad_norm": 1.5078125,
-      "learning_rate": 4.673471437148973e-06,
-      "long_answer_loss": 0.0895,
-      "loss": 0.0745,
-      "short_answer_loss": NaN,
-      "step": 895,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0645,
-      "grad_norm": 1.3828125,
-      "learning_rate": 4.64793934181431e-06,
-      "long_answer_loss": 0.0645,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 896,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0683,
-      "grad_norm": 1.34375,
-      "learning_rate": 4.6224612434640575e-06,
-      "long_answer_loss": 0.0683,
-      "loss": 0.0634,
-      "short_answer_loss": NaN,
-      "step": 897,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0634,
-      "grad_norm": 1.4453125,
-      "learning_rate": 4.597037317305788e-06,
-      "long_answer_loss": 0.0634,
-      "loss": 0.0654,
-      "short_answer_loss": NaN,
-      "step": 898,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.45,
-      "full_loss": 0.0683,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.571667738174547e-06,
-      "long_answer_loss": 0.0683,
-      "loss": 0.0709,
-      "short_answer_loss": NaN,
-      "step": 899,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0723,
-      "grad_norm": 1.5625,
-      "learning_rate": 4.546352680531639e-06,
-      "long_answer_loss": 0.0723,
-      "loss": 0.0689,
-      "short_answer_loss": NaN,
-      "step": 900,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0885,
-      "grad_norm": 1.5078125,
-      "learning_rate": 4.521092318463439e-06,
-      "long_answer_loss": 0.0885,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 901,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0772,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.495886825680192e-06,
-      "long_answer_loss": 0.0772,
-      "loss": 0.0681,
-      "short_answer_loss": NaN,
-      "step": 902,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0694,
-      "grad_norm": 1.6171875,
-      "learning_rate": 4.470736375514818e-06,
-      "long_answer_loss": 0.0694,
-      "loss": 0.0776,
-      "short_answer_loss": NaN,
-      "step": 903,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0899,
-      "grad_norm": 1.5,
-      "learning_rate": 4.445641140921721e-06,
-      "long_answer_loss": 0.0899,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 904,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0637,
-      "grad_norm": 1.5,
-      "learning_rate": 4.420601294475595e-06,
-      "long_answer_loss": 0.0637,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 905,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.46,
-      "full_loss": 0.0602,
-      "grad_norm": 1.375,
-      "learning_rate": 4.395617008370248e-06,
-      "long_answer_loss": 0.0602,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 906,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0693,
-      "grad_norm": 1.4921875,
-      "learning_rate": 4.370688454417405e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 907,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0855,
-      "grad_norm": 1.4921875,
-      "learning_rate": 4.345815804045539e-06,
-      "long_answer_loss": 0.0855,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 908,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0583,
-      "grad_norm": 1.3828125,
-      "learning_rate": 4.320999228298678e-06,
-      "long_answer_loss": 0.0583,
-      "loss": 0.064,
-      "short_answer_loss": NaN,
-      "step": 909,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0899,
-      "grad_norm": 1.4765625,
-      "learning_rate": 4.2962388978352435e-06,
-      "long_answer_loss": 0.0899,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 910,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0773,
-      "grad_norm": 1.515625,
-      "learning_rate": 4.271534982926864e-06,
-      "long_answer_loss": 0.0773,
-      "loss": 0.0704,
-      "short_answer_loss": NaN,
-      "step": 911,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.47,
-      "full_loss": 0.0494,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.246887653457216e-06,
-      "long_answer_loss": 0.0494,
-      "loss": 0.0607,
-      "short_answer_loss": NaN,
-      "step": 912,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0653,
-      "grad_norm": 1.4296875,
-      "learning_rate": 4.222297078920845e-06,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 913,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0765,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.197763428422005e-06,
-      "long_answer_loss": 0.0765,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 914,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0643,
-      "grad_norm": 1.4453125,
-      "learning_rate": 4.173286870673498e-06,
-      "long_answer_loss": 0.0643,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 915,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0814,
-      "grad_norm": 1.5078125,
-      "learning_rate": 4.148867573995511e-06,
-      "long_answer_loss": 0.0814,
-      "loss": 0.0765,
-      "short_answer_loss": NaN,
-      "step": 916,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0628,
-      "grad_norm": 1.4453125,
-      "learning_rate": 4.124505706314455e-06,
-      "long_answer_loss": 0.0628,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 917,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.48,
-      "full_loss": 0.0651,
-      "grad_norm": 1.625,
-      "learning_rate": 4.100201435161817e-06,
-      "long_answer_loss": 0.0651,
-      "loss": 0.0749,
-      "short_answer_loss": NaN,
-      "step": 918,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.089,
-      "grad_norm": 1.375,
-      "learning_rate": 4.0759549276730025e-06,
-      "long_answer_loss": 0.089,
-      "loss": 0.0727,
-      "short_answer_loss": NaN,
-      "step": 919,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0845,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.051766350586187e-06,
-      "long_answer_loss": 0.0845,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 920,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0693,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.027635870241178e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0664,
-      "short_answer_loss": NaN,
-      "step": 921,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0697,
-      "grad_norm": 1.484375,
-      "learning_rate": 4.003563652578258e-06,
-      "long_answer_loss": 0.0697,
-      "loss": 0.07,
-      "short_answer_loss": NaN,
-      "step": 922,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0664,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.9795498631370515e-06,
-      "long_answer_loss": 0.0664,
-      "loss": 0.0751,
-      "short_answer_loss": NaN,
-      "step": 923,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.49,
-      "full_loss": 0.0788,
-      "grad_norm": 1.4921875,
-      "learning_rate": 3.9555946670553774e-06,
-      "long_answer_loss": 0.0788,
-      "loss": 0.0727,
-      "short_answer_loss": NaN,
-      "step": 924,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0861,
-      "grad_norm": 1.5,
-      "learning_rate": 3.931698229068131e-06,
-      "long_answer_loss": 0.0861,
-      "loss": 0.0738,
-      "short_answer_loss": NaN,
-      "step": 925,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0731,
-      "grad_norm": 1.4609375,
-      "learning_rate": 3.907860713506132e-06,
-      "long_answer_loss": 0.0731,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 926,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0647,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.884082284295008e-06,
-      "long_answer_loss": 0.0647,
-      "loss": 0.072,
-      "short_answer_loss": NaN,
-      "step": 927,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0623,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.860363104954059e-06,
-      "long_answer_loss": 0.0623,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 928,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0813,
-      "grad_norm": 1.4140625,
-      "learning_rate": 3.836703338595138e-06,
-      "long_answer_loss": 0.0813,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 929,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.5,
-      "full_loss": 0.0863,
-      "grad_norm": 1.5625,
-      "learning_rate": 3.813103147921526e-06,
-      "long_answer_loss": 0.0863,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 930,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0466,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.7895626952268155e-06,
-      "long_answer_loss": 0.0466,
-      "loss": 0.0638,
-      "short_answer_loss": NaN,
-      "step": 931,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0831,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.766082142393791e-06,
-      "long_answer_loss": 0.0831,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 932,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0468,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.7426616508933214e-06,
-      "long_answer_loss": 0.0468,
-      "loss": 0.0678,
-      "short_answer_loss": NaN,
-      "step": 933,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0818,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.7193013817832454e-06,
-      "long_answer_loss": 0.0818,
-      "loss": 0.0737,
-      "short_answer_loss": NaN,
-      "step": 934,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0733,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.696001495707263e-06,
-      "long_answer_loss": 0.0733,
-      "loss": 0.0689,
-      "short_answer_loss": NaN,
-      "step": 935,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.067,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.672762152893834e-06,
-      "long_answer_loss": 0.067,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 936,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.51,
-      "full_loss": 0.0583,
-      "grad_norm": 1.53125,
-      "learning_rate": 3.6495835131550748e-06,
-      "long_answer_loss": 0.0583,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 937,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0655,
-      "grad_norm": 1.4609375,
-      "learning_rate": 3.6264657358856604e-06,
-      "long_answer_loss": 0.0655,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 938,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0497,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.603408980061726e-06,
-      "long_answer_loss": 0.0497,
-      "loss": 0.0665,
-      "short_answer_loss": NaN,
-      "step": 939,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0841,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.5804134042397743e-06,
-      "long_answer_loss": 0.0841,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 940,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.0685,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.5574791665555882e-06,
-      "long_answer_loss": 0.0685,
-      "loss": 0.0672,
-      "short_answer_loss": NaN,
-      "step": 941,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.071,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.5346064247231387e-06,
-      "long_answer_loss": 0.071,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 942,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.52,
-      "full_loss": 0.059,
-      "grad_norm": 1.4609375,
-      "learning_rate": 3.511795336033505e-06,
-      "long_answer_loss": 0.059,
-      "loss": 0.0691,
-      "short_answer_loss": NaN,
-      "step": 943,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0613,
-      "grad_norm": 1.375,
-      "learning_rate": 3.489046057353787e-06,
-      "long_answer_loss": 0.0613,
-      "loss": 0.0661,
-      "short_answer_loss": NaN,
-      "step": 944,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0751,
-      "grad_norm": 1.59375,
-      "learning_rate": 3.466358745126033e-06,
-      "long_answer_loss": 0.0751,
-      "loss": 0.071,
-      "short_answer_loss": NaN,
-      "step": 945,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0732,
-      "grad_norm": 1.5546875,
-      "learning_rate": 3.4437335553661605e-06,
-      "long_answer_loss": 0.0732,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 946,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0718,
-      "grad_norm": 1.4765625,
-      "learning_rate": 3.421170643662884e-06,
-      "long_answer_loss": 0.0718,
-      "loss": 0.0737,
-      "short_answer_loss": NaN,
-      "step": 947,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.0718,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.3986701651766426e-06,
-      "long_answer_loss": 0.0718,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 948,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.53,
-      "full_loss": 0.071,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.3762322746385383e-06,
-      "long_answer_loss": 0.071,
-      "loss": 0.0695,
-      "short_answer_loss": NaN,
-      "step": 949,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0776,
-      "grad_norm": 1.53125,
-      "learning_rate": 3.353857126349265e-06,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0751,
-      "short_answer_loss": NaN,
-      "step": 950,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0624,
-      "grad_norm": 1.5625,
-      "learning_rate": 3.3315448741780566e-06,
-      "long_answer_loss": 0.0624,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 951,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0799,
-      "grad_norm": 1.359375,
-      "learning_rate": 3.309295671561617e-06,
-      "long_answer_loss": 0.0799,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 952,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.079,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.287109671503079e-06,
-      "long_answer_loss": 0.079,
-      "loss": 0.0684,
-      "short_answer_loss": NaN,
-      "step": 953,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0607,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.2649870265709314e-06,
-      "long_answer_loss": 0.0607,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 954,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.54,
-      "full_loss": 0.0714,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.2429278888980034e-06,
-      "long_answer_loss": 0.0714,
-      "loss": 0.0761,
-      "short_answer_loss": NaN,
-      "step": 955,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0825,
-      "grad_norm": 1.515625,
-      "learning_rate": 3.220932410180383e-06,
-      "long_answer_loss": 0.0825,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 956,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.088,
-      "grad_norm": 1.5,
-      "learning_rate": 3.1990007416763904e-06,
-      "long_answer_loss": 0.088,
-      "loss": 0.0716,
-      "short_answer_loss": NaN,
-      "step": 957,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0535,
-      "grad_norm": 1.375,
-      "learning_rate": 3.1771330342055387e-06,
-      "long_answer_loss": 0.0535,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 958,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0861,
-      "grad_norm": 1.46875,
-      "learning_rate": 3.1553294381474946e-06,
-      "long_answer_loss": 0.0861,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 959,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.0687,
-      "grad_norm": 1.3984375,
-      "learning_rate": 3.133590103441042e-06,
-      "long_answer_loss": 0.0687,
-      "loss": 0.069,
-      "short_answer_loss": NaN,
-      "step": 960,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.55,
-      "full_loss": 0.066,
-      "grad_norm": 1.3515625,
-      "learning_rate": 3.1119151795830567e-06,
-      "long_answer_loss": 0.066,
-      "loss": 0.0673,
-      "short_answer_loss": NaN,
-      "step": 961,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0574,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.0903048156274707e-06,
-      "long_answer_loss": 0.0574,
-      "loss": 0.0683,
-      "short_answer_loss": NaN,
-      "step": 962,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.068,
-      "grad_norm": 1.4609375,
-      "learning_rate": 3.0687591601842524e-06,
-      "long_answer_loss": 0.068,
-      "loss": 0.0683,
-      "short_answer_loss": NaN,
-      "step": 963,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.056,
-      "grad_norm": 1.375,
-      "learning_rate": 3.047278361418382e-06,
-      "long_answer_loss": 0.056,
-      "loss": 0.0713,
-      "short_answer_loss": NaN,
-      "step": 964,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0583,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.0258625670488373e-06,
-      "long_answer_loss": 0.0583,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 965,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0657,
-      "grad_norm": 1.515625,
-      "learning_rate": 3.0045119243475696e-06,
-      "long_answer_loss": 0.0657,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 966,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.56,
-      "full_loss": 0.0676,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.9832265801385e-06,
-      "long_answer_loss": 0.0676,
-      "loss": 0.0658,
-      "short_answer_loss": NaN,
-      "step": 967,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0581,
-      "grad_norm": 1.640625,
-      "learning_rate": 2.962006680796503e-06,
-      "long_answer_loss": 0.0581,
-      "loss": 0.0727,
-      "short_answer_loss": NaN,
-      "step": 968,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.093,
-      "grad_norm": 1.375,
-      "learning_rate": 2.940852372246404e-06,
-      "long_answer_loss": 0.093,
-      "loss": 0.062,
-      "short_answer_loss": NaN,
-      "step": 969,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0712,
-      "grad_norm": 1.546875,
-      "learning_rate": 2.9197637999619733e-06,
-      "long_answer_loss": 0.0712,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 970,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0704,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.898741108964925e-06,
-      "long_answer_loss": 0.0704,
-      "loss": 0.0712,
-      "short_answer_loss": NaN,
-      "step": 971,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0738,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.877784443823926e-06,
-      "long_answer_loss": 0.0738,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 972,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0691,
-      "grad_norm": 1.5625,
-      "learning_rate": 2.856893948653591e-06,
-      "long_answer_loss": 0.0691,
-      "loss": 0.0727,
-      "short_answer_loss": NaN,
-      "step": 973,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.57,
-      "full_loss": 0.0732,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.836069767113503e-06,
-      "long_answer_loss": 0.0732,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 974,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0636,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.8153120424072156e-06,
-      "long_answer_loss": 0.0636,
-      "loss": 0.0707,
-      "short_answer_loss": NaN,
-      "step": 975,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.087,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.794620917281278e-06,
-      "long_answer_loss": 0.087,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 976,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0716,
-      "grad_norm": 1.3046875,
-      "learning_rate": 2.773996534024241e-06,
-      "long_answer_loss": 0.0716,
-      "loss": 0.0605,
-      "short_answer_loss": NaN,
-      "step": 977,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0541,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.753439034465695e-06,
-      "long_answer_loss": 0.0541,
-      "loss": 0.0657,
-      "short_answer_loss": NaN,
-      "step": 978,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0486,
-      "grad_norm": 1.3671875,
-      "learning_rate": 2.732948559975271e-06,
-      "long_answer_loss": 0.0486,
-      "loss": 0.062,
-      "short_answer_loss": NaN,
-      "step": 979,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.58,
-      "full_loss": 0.0765,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.7125252514616966e-06,
-      "long_answer_loss": 0.0765,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 980,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0601,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.692169249371804e-06,
-      "long_answer_loss": 0.0601,
-      "loss": 0.0653,
-      "short_answer_loss": NaN,
-      "step": 981,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0817,
-      "grad_norm": 1.375,
-      "learning_rate": 2.6718806936895796e-06,
-      "long_answer_loss": 0.0817,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 982,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.066,
-      "grad_norm": 1.4453125,
-      "learning_rate": 2.651659723935189e-06,
-      "long_answer_loss": 0.066,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 983,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0578,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.6315064791640296e-06,
-      "long_answer_loss": 0.0578,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 984,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.0679,
-      "grad_norm": 1.6796875,
-      "learning_rate": 2.6114210979657606e-06,
-      "long_answer_loss": 0.0679,
-      "loss": 0.0746,
-      "short_answer_loss": NaN,
-      "step": 985,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.59,
-      "full_loss": 0.064,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.5914037184633656e-06,
-      "long_answer_loss": 0.064,
-      "loss": 0.064,
-      "short_answer_loss": NaN,
-      "step": 986,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0693,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.571454478312185e-06,
-      "long_answer_loss": 0.0693,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 987,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0723,
-      "grad_norm": 1.46875,
-      "learning_rate": 2.5515735146989933e-06,
-      "long_answer_loss": 0.0723,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 988,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0792,
-      "grad_norm": 1.5703125,
-      "learning_rate": 2.531760964341029e-06,
-      "long_answer_loss": 0.0792,
-      "loss": 0.0716,
-      "short_answer_loss": NaN,
-      "step": 989,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0577,
-      "grad_norm": 1.5234375,
-      "learning_rate": 2.5120169634850713e-06,
-      "long_answer_loss": 0.0577,
-      "loss": 0.0652,
-      "short_answer_loss": NaN,
-      "step": 990,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0653,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.4923416479064987e-06,
-      "long_answer_loss": 0.0653,
-      "loss": 0.064,
-      "short_answer_loss": NaN,
-      "step": 991,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.6,
-      "full_loss": 0.0761,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.4727351529083536e-06,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 992,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0722,
-      "grad_norm": 1.34375,
-      "learning_rate": 2.4531976133204184e-06,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0642,
-      "short_answer_loss": NaN,
-      "step": 993,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0715,
-      "grad_norm": 1.34375,
-      "learning_rate": 2.4337291634982757e-06,
-      "long_answer_loss": 0.0715,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 994,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0635,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.4143299373224015e-06,
-      "long_answer_loss": 0.0635,
-      "loss": 0.0723,
-      "short_answer_loss": NaN,
-      "step": 995,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0864,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.3950000681972284e-06,
-      "long_answer_loss": 0.0864,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 996,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0716,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.3757396890502382e-06,
-      "long_answer_loss": 0.0716,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 997,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.61,
-      "full_loss": 0.0498,
-      "grad_norm": 1.3828125,
-      "learning_rate": 2.3565489323310402e-06,
-      "long_answer_loss": 0.0498,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 998,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0565,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.3374279300104733e-06,
-      "long_answer_loss": 0.0565,
-      "loss": 0.0625,
-      "short_answer_loss": NaN,
-      "step": 999,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0556,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.31837681357968e-06,
-      "long_answer_loss": 0.0556,
-      "loss": 0.0653,
-      "short_answer_loss": NaN,
-      "step": 1000,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0831,
-      "grad_norm": 1.5078125,
-      "learning_rate": 2.2993957140492197e-06,
-      "long_answer_loss": 0.0831,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 1001,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0598,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.2804847619481552e-06,
-      "long_answer_loss": 0.0598,
-      "loss": 0.0646,
-      "short_answer_loss": NaN,
-      "step": 1002,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0641,
-      "grad_norm": 1.6171875,
-      "learning_rate": 2.2616440873231655e-06,
-      "long_answer_loss": 0.0641,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 1003,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0806,
-      "grad_norm": 1.5078125,
-      "learning_rate": 2.2428738197376397e-06,
-      "long_answer_loss": 0.0806,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 1004,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.62,
-      "full_loss": 0.0572,
-      "grad_norm": 1.3984375,
-      "learning_rate": 2.224174088270796e-06,
-      "long_answer_loss": 0.0572,
-      "loss": 0.0665,
-      "short_answer_loss": NaN,
-      "step": 1005,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0618,
-      "grad_norm": 1.53125,
-      "learning_rate": 2.20554502151679e-06,
-      "long_answer_loss": 0.0618,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1006,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0638,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.1869867475838317e-06,
-      "long_answer_loss": 0.0638,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 1007,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0662,
-      "grad_norm": 1.390625,
-      "learning_rate": 2.1684993940933013e-06,
-      "long_answer_loss": 0.0662,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 1008,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0761,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.150083088178875e-06,
-      "long_answer_loss": 0.0761,
-      "loss": 0.0658,
-      "short_answer_loss": NaN,
-      "step": 1009,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0672,
-      "grad_norm": 1.546875,
-      "learning_rate": 2.131737956485652e-06,
-      "long_answer_loss": 0.0672,
-      "loss": 0.0746,
-      "short_answer_loss": NaN,
-      "step": 1010,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.63,
-      "full_loss": 0.0695,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.113464125169276e-06,
-      "long_answer_loss": 0.0695,
-      "loss": 0.0657,
-      "short_answer_loss": NaN,
-      "step": 1011,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0669,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.0952617198950765e-06,
-      "long_answer_loss": 0.0669,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 1012,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0679,
-      "grad_norm": 1.4921875,
-      "learning_rate": 2.0771308658372015e-06,
-      "long_answer_loss": 0.0679,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1013,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0656,
-      "grad_norm": 1.5,
-      "learning_rate": 2.059071687677755e-06,
-      "long_answer_loss": 0.0656,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 1014,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0591,
-      "grad_norm": 1.515625,
-      "learning_rate": 2.0410843096059394e-06,
-      "long_answer_loss": 0.0591,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 1015,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0654,
-      "grad_norm": 1.5625,
-      "learning_rate": 2.0231688553172064e-06,
-      "long_answer_loss": 0.0654,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1016,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.64,
-      "full_loss": 0.0745,
-      "grad_norm": 1.5234375,
-      "learning_rate": 2.0053254480123977e-06,
-      "long_answer_loss": 0.0745,
-      "loss": 0.0734,
-      "short_answer_loss": NaN,
-      "step": 1017,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0657,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.9875542103969094e-06,
-      "long_answer_loss": 0.0657,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 1018,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0704,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.969855264679836e-06,
-      "long_answer_loss": 0.0704,
-      "loss": 0.0684,
-      "short_answer_loss": NaN,
-      "step": 1019,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.07,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.9522287325731357e-06,
-      "long_answer_loss": 0.07,
-      "loss": 0.0666,
-      "short_answer_loss": NaN,
-      "step": 1020,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.061,
-      "grad_norm": 1.3984375,
-      "learning_rate": 1.934674735290802e-06,
-      "long_answer_loss": 0.061,
-      "loss": 0.0726,
-      "short_answer_loss": NaN,
-      "step": 1021,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0564,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.91719339354801e-06,
-      "long_answer_loss": 0.0564,
-      "loss": 0.0674,
-      "short_answer_loss": NaN,
-      "step": 1022,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.65,
-      "full_loss": 0.0543,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.8997848275603067e-06,
-      "long_answer_loss": 0.0543,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1023,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0813,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.8824491570427676e-06,
-      "long_answer_loss": 0.0813,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 1024,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0644,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.8651865012091888e-06,
-      "long_answer_loss": 0.0644,
-      "loss": 0.0631,
-      "short_answer_loss": NaN,
-      "step": 1025,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0845,
-      "grad_norm": 1.3671875,
-      "learning_rate": 1.8479969787712533e-06,
-      "long_answer_loss": 0.0845,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 1026,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0604,
-      "grad_norm": 1.6953125,
-      "learning_rate": 1.830880707937725e-06,
-      "long_answer_loss": 0.0604,
-      "loss": 0.0776,
-      "short_answer_loss": NaN,
-      "step": 1027,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0618,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.8138378064136318e-06,
-      "long_answer_loss": 0.0618,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 1028,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.66,
-      "full_loss": 0.0637,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.796868391399452e-06,
-      "long_answer_loss": 0.0637,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 1029,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0725,
-      "grad_norm": 1.5625,
-      "learning_rate": 1.7799725795903193e-06,
-      "long_answer_loss": 0.0725,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1030,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0685,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.7631504871752066e-06,
-      "long_answer_loss": 0.0685,
-      "loss": 0.0648,
-      "short_answer_loss": NaN,
-      "step": 1031,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0777,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.7464022298361374e-06,
-      "long_answer_loss": 0.0777,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 1032,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0717,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.7297279227473874e-06,
-      "long_answer_loss": 0.0717,
-      "loss": 0.0673,
-      "short_answer_loss": NaN,
-      "step": 1033,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0736,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.7131276805746902e-06,
-      "long_answer_loss": 0.0736,
-      "loss": 0.072,
-      "short_answer_loss": NaN,
-      "step": 1034,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.67,
-      "full_loss": 0.0637,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.6966016174744499e-06,
-      "long_answer_loss": 0.0637,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 1035,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0729,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.68014984709296e-06,
-      "long_answer_loss": 0.0729,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 1036,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0574,
-      "grad_norm": 1.3828125,
-      "learning_rate": 1.6637724825656147e-06,
-      "long_answer_loss": 0.0574,
-      "loss": 0.0669,
-      "short_answer_loss": NaN,
-      "step": 1037,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0873,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.6474696365161358e-06,
-      "long_answer_loss": 0.0873,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 1038,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.064,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.6312414210557972e-06,
-      "long_answer_loss": 0.064,
-      "loss": 0.0688,
-      "short_answer_loss": NaN,
-      "step": 1039,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0635,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.615087947782655e-06,
-      "long_answer_loss": 0.0635,
-      "loss": 0.0644,
-      "short_answer_loss": NaN,
-      "step": 1040,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.073,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.5990093277807775e-06,
-      "long_answer_loss": 0.073,
-      "loss": 0.0657,
-      "short_answer_loss": NaN,
-      "step": 1041,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.68,
-      "full_loss": 0.0586,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.583005671619482e-06,
-      "long_answer_loss": 0.0586,
-      "loss": 0.0662,
-      "short_answer_loss": NaN,
-      "step": 1042,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0633,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.5670770893525768e-06,
-      "long_answer_loss": 0.0633,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 1043,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0764,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.5512236905176018e-06,
-      "long_answer_loss": 0.0764,
-      "loss": 0.0657,
-      "short_answer_loss": NaN,
-      "step": 1044,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0818,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.5354455841350756e-06,
-      "long_answer_loss": 0.0818,
-      "loss": 0.0619,
-      "short_answer_loss": NaN,
-      "step": 1045,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0697,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.5197428787077472e-06,
-      "long_answer_loss": 0.0697,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 1046,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0705,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.5041156822198492e-06,
-      "long_answer_loss": 0.0705,
-      "loss": 0.0626,
-      "short_answer_loss": NaN,
-      "step": 1047,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.69,
-      "full_loss": 0.0695,
-      "grad_norm": 1.5,
-      "learning_rate": 1.4885641021363541e-06,
-      "long_answer_loss": 0.0695,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 1048,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0676,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.4730882454022362e-06,
-      "long_answer_loss": 0.0676,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1049,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0787,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.457688218441737e-06,
-      "long_answer_loss": 0.0787,
-      "loss": 0.0769,
-      "short_answer_loss": NaN,
-      "step": 1050,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0595,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.442364127157632e-06,
-      "long_answer_loss": 0.0595,
-      "loss": 0.0766,
-      "short_answer_loss": NaN,
-      "step": 1051,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0641,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.4271160769305014e-06,
-      "long_answer_loss": 0.0641,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 1052,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0668,
-      "grad_norm": 1.359375,
-      "learning_rate": 1.4119441726180085e-06,
-      "long_answer_loss": 0.0668,
-      "loss": 0.064,
-      "short_answer_loss": NaN,
-      "step": 1053,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.7,
-      "full_loss": 0.0754,
-      "grad_norm": 1.3203125,
-      "learning_rate": 1.396848518554178e-06,
-      "long_answer_loss": 0.0754,
-      "loss": 0.0658,
-      "short_answer_loss": NaN,
-      "step": 1054,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0611,
-      "grad_norm": 1.4609375,
-      "learning_rate": 1.3818292185486749e-06,
-      "long_answer_loss": 0.0611,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1055,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0653,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.366886375886095e-06,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0735,
-      "short_answer_loss": NaN,
-      "step": 1056,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0558,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.3520200933252542e-06,
-      "long_answer_loss": 0.0558,
-      "loss": 0.0654,
-      "short_answer_loss": NaN,
-      "step": 1057,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0661,
-      "grad_norm": 1.5703125,
-      "learning_rate": 1.337230473098476e-06,
-      "long_answer_loss": 0.0661,
-      "loss": 0.07,
-      "short_answer_loss": NaN,
-      "step": 1058,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0689,
-      "grad_norm": 1.609375,
-      "learning_rate": 1.322517616910897e-06,
-      "long_answer_loss": 0.0689,
-      "loss": 0.0716,
-      "short_answer_loss": NaN,
-      "step": 1059,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.71,
-      "full_loss": 0.0685,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.3078816259397635e-06,
-      "long_answer_loss": 0.0685,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1060,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0617,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.2933226008337324e-06,
-      "long_answer_loss": 0.0617,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1061,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0637,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.2788406417121867e-06,
-      "long_answer_loss": 0.0637,
-      "loss": 0.0643,
-      "short_answer_loss": NaN,
-      "step": 1062,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0675,
-      "grad_norm": 1.5,
-      "learning_rate": 1.2644358481645399e-06,
-      "long_answer_loss": 0.0675,
-      "loss": 0.0733,
-      "short_answer_loss": NaN,
-      "step": 1063,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0769,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.2501083192495544e-06,
-      "long_answer_loss": 0.0769,
-      "loss": 0.0676,
-      "short_answer_loss": NaN,
-      "step": 1064,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0555,
-      "grad_norm": 1.46875,
-      "learning_rate": 1.2358581534946594e-06,
-      "long_answer_loss": 0.0555,
-      "loss": 0.0682,
-      "short_answer_loss": NaN,
-      "step": 1065,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.72,
-      "full_loss": 0.0875,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.2216854488952753e-06,
-      "long_answer_loss": 0.0875,
-      "loss": 0.0691,
-      "short_answer_loss": NaN,
-      "step": 1066,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0669,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.2075903029141384e-06,
-      "long_answer_loss": 0.0669,
-      "loss": 0.065,
-      "short_answer_loss": NaN,
-      "step": 1067,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0809,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.193572812480627e-06,
-      "long_answer_loss": 0.0809,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 1068,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0661,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.1796330739901024e-06,
-      "long_answer_loss": 0.0661,
-      "loss": 0.0676,
-      "short_answer_loss": NaN,
-      "step": 1069,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0572,
-      "grad_norm": 1.484375,
-      "learning_rate": 1.1657711833032394e-06,
-      "long_answer_loss": 0.0572,
-      "loss": 0.0651,
-      "short_answer_loss": NaN,
-      "step": 1070,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0708,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.1519872357453734e-06,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0682,
-      "short_answer_loss": NaN,
-      "step": 1071,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0714,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.1382813261058349e-06,
-      "long_answer_loss": 0.0714,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 1072,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.73,
-      "full_loss": 0.0728,
-      "grad_norm": 1.375,
-      "learning_rate": 1.124653548637311e-06,
-      "long_answer_loss": 0.0728,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 1073,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0528,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.111103997055185e-06,
-      "long_answer_loss": 0.0528,
-      "loss": 0.0657,
-      "short_answer_loss": NaN,
-      "step": 1074,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0611,
-      "grad_norm": 1.375,
-      "learning_rate": 1.0976327645368975e-06,
-      "long_answer_loss": 0.0611,
-      "loss": 0.0662,
-      "short_answer_loss": NaN,
-      "step": 1075,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0783,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.0842399437213103e-06,
-      "long_answer_loss": 0.0783,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 1076,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0896,
-      "grad_norm": 1.5859375,
-      "learning_rate": 1.0709256267080566e-06,
-      "long_answer_loss": 0.0896,
-      "loss": 0.0732,
-      "short_answer_loss": NaN,
-      "step": 1077,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0789,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.0576899050569204e-06,
-      "long_answer_loss": 0.0789,
-      "loss": 0.0686,
-      "short_answer_loss": NaN,
-      "step": 1078,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.74,
-      "full_loss": 0.0569,
-      "grad_norm": 1.5,
-      "learning_rate": 1.0445328697872015e-06,
-      "long_answer_loss": 0.0569,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 1079,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0676,
-      "grad_norm": 1.5234375,
-      "learning_rate": 1.0314546113770876e-06,
-      "long_answer_loss": 0.0676,
-      "loss": 0.0672,
-      "short_answer_loss": NaN,
-      "step": 1080,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0694,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.018455219763037e-06,
-      "long_answer_loss": 0.0694,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 1081,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0741,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.0055347843391557e-06,
-      "long_answer_loss": 0.0741,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1082,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0622,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.92693393956584e-07,
-      "long_answer_loss": 0.0622,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 1083,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0568,
-      "grad_norm": 1.4140625,
-      "learning_rate": 9.79931136922889e-07,
-      "long_answer_loss": 0.0568,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1084,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.75,
-      "full_loss": 0.0558,
-      "grad_norm": 1.3671875,
-      "learning_rate": 9.672481010014486e-07,
-      "long_answer_loss": 0.0558,
-      "loss": 0.0639,
-      "short_answer_loss": NaN,
-      "step": 1085,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0688,
-      "grad_norm": 1.4609375,
-      "learning_rate": 9.54644373410861e-07,
-      "long_answer_loss": 0.0688,
-      "loss": 0.0715,
-      "short_answer_loss": NaN,
-      "step": 1086,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.1051,
-      "grad_norm": 1.546875,
-      "learning_rate": 9.421200408243277e-07,
-      "long_answer_loss": 0.1051,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 1087,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0607,
-      "grad_norm": 1.421875,
-      "learning_rate": 9.296751893690808e-07,
-      "long_answer_loss": 0.0607,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 1088,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0595,
-      "grad_norm": 1.4765625,
-      "learning_rate": 9.173099046257655e-07,
-      "long_answer_loss": 0.0595,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 1089,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0611,
-      "grad_norm": 1.53125,
-      "learning_rate": 9.050242716278676e-07,
-      "long_answer_loss": 0.0611,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 1090,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.76,
-      "full_loss": 0.0559,
-      "grad_norm": 1.4765625,
-      "learning_rate": 8.928183748611263e-07,
-      "long_answer_loss": 0.0559,
-      "loss": 0.0701,
-      "short_answer_loss": NaN,
-      "step": 1091,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0727,
-      "grad_norm": 1.3984375,
-      "learning_rate": 8.806922982629473e-07,
-      "long_answer_loss": 0.0727,
-      "loss": 0.0642,
-      "short_answer_loss": NaN,
-      "step": 1092,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0722,
-      "grad_norm": 1.5390625,
-      "learning_rate": 8.686461252218323e-07,
-      "long_answer_loss": 0.0722,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 1093,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0591,
-      "grad_norm": 1.3671875,
-      "learning_rate": 8.566799385768015e-07,
-      "long_answer_loss": 0.0591,
-      "loss": 0.0672,
-      "short_answer_loss": NaN,
-      "step": 1094,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0665,
-      "grad_norm": 1.3828125,
-      "learning_rate": 8.447938206168279e-07,
-      "long_answer_loss": 0.0665,
-      "loss": 0.0645,
-      "short_answer_loss": NaN,
-      "step": 1095,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.0767,
-      "grad_norm": 1.4296875,
-      "learning_rate": 8.329878530802665e-07,
-      "long_answer_loss": 0.0767,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 1096,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.77,
-      "full_loss": 0.058,
-      "grad_norm": 1.375,
-      "learning_rate": 8.21262117154295e-07,
-      "long_answer_loss": 0.058,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 1097,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0668,
-      "grad_norm": 1.4140625,
-      "learning_rate": 8.096166934743549e-07,
-      "long_answer_loss": 0.0668,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1098,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0738,
-      "grad_norm": 1.625,
-      "learning_rate": 7.980516621235973e-07,
-      "long_answer_loss": 0.0738,
-      "loss": 0.0747,
-      "short_answer_loss": NaN,
-      "step": 1099,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0578,
-      "grad_norm": 1.421875,
-      "learning_rate": 7.865671026323323e-07,
-      "long_answer_loss": 0.0578,
-      "loss": 0.0638,
-      "short_answer_loss": NaN,
-      "step": 1100,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0657,
-      "grad_norm": 1.40625,
-      "learning_rate": 7.751630939774823e-07,
-      "long_answer_loss": 0.0657,
-      "loss": 0.0627,
-      "short_answer_loss": NaN,
-      "step": 1101,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0713,
-      "grad_norm": 1.453125,
-      "learning_rate": 7.638397145820361e-07,
-      "long_answer_loss": 0.0713,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1102,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0662,
-      "grad_norm": 1.4921875,
-      "learning_rate": 7.525970423145165e-07,
-      "long_answer_loss": 0.0662,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 1103,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.78,
-      "full_loss": 0.0626,
-      "grad_norm": 1.515625,
-      "learning_rate": 7.414351544884332e-07,
-      "long_answer_loss": 0.0626,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 1104,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0612,
-      "grad_norm": 1.5546875,
-      "learning_rate": 7.303541278617654e-07,
-      "long_answer_loss": 0.0612,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 1105,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0803,
-      "grad_norm": 1.46875,
-      "learning_rate": 7.193540386364203e-07,
-      "long_answer_loss": 0.0803,
-      "loss": 0.0755,
-      "short_answer_loss": NaN,
-      "step": 1106,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.1016,
-      "grad_norm": 1.4765625,
-      "learning_rate": 7.084349624577213e-07,
-      "long_answer_loss": 0.1016,
-      "loss": 0.0779,
-      "short_answer_loss": NaN,
-      "step": 1107,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0583,
-      "grad_norm": 1.3828125,
-      "learning_rate": 6.975969744138791e-07,
-      "long_answer_loss": 0.0583,
-      "loss": 0.0642,
-      "short_answer_loss": NaN,
-      "step": 1108,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0548,
-      "grad_norm": 1.4921875,
-      "learning_rate": 6.868401490354767e-07,
-      "long_answer_loss": 0.0548,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 1109,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.79,
-      "full_loss": 0.0653,
-      "grad_norm": 1.53125,
-      "learning_rate": 6.761645602949618e-07,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 1110,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0604,
-      "grad_norm": 1.4140625,
-      "learning_rate": 6.655702816061316e-07,
-      "long_answer_loss": 0.0604,
-      "loss": 0.0645,
-      "short_answer_loss": NaN,
-      "step": 1111,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0587,
-      "grad_norm": 1.46875,
-      "learning_rate": 6.55057385823632e-07,
-      "long_answer_loss": 0.0587,
-      "loss": 0.0689,
-      "short_answer_loss": NaN,
-      "step": 1112,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0708,
-      "grad_norm": 1.3984375,
-      "learning_rate": 6.446259452424566e-07,
-      "long_answer_loss": 0.0708,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 1113,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0642,
-      "grad_norm": 1.5546875,
-      "learning_rate": 6.342760315974485e-07,
-      "long_answer_loss": 0.0642,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1114,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0649,
-      "grad_norm": 1.46875,
-      "learning_rate": 6.240077160628063e-07,
-      "long_answer_loss": 0.0649,
-      "loss": 0.0692,
-      "short_answer_loss": NaN,
-      "step": 1115,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.8,
-      "full_loss": 0.0659,
-      "grad_norm": 1.5859375,
-      "learning_rate": 6.138210692515939e-07,
-      "long_answer_loss": 0.0659,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 1116,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0616,
-      "grad_norm": 1.3984375,
-      "learning_rate": 6.037161612152606e-07,
-      "long_answer_loss": 0.0616,
-      "loss": 0.0654,
-      "short_answer_loss": NaN,
-      "step": 1117,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0791,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.936930614431499e-07,
-      "long_answer_loss": 0.0791,
-      "loss": 0.0715,
-      "short_answer_loss": NaN,
-      "step": 1118,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0768,
-      "grad_norm": 1.6328125,
-      "learning_rate": 5.837518388620317e-07,
-      "long_answer_loss": 0.0768,
-      "loss": 0.0759,
-      "short_answer_loss": NaN,
-      "step": 1119,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0762,
-      "grad_norm": 1.3515625,
-      "learning_rate": 5.738925618356206e-07,
-      "long_answer_loss": 0.0762,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 1120,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0957,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.641152981641084e-07,
-      "long_answer_loss": 0.0957,
-      "loss": 0.0706,
-      "short_answer_loss": NaN,
-      "step": 1121,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.81,
-      "full_loss": 0.0673,
-      "grad_norm": 1.4296875,
-      "learning_rate": 5.544201150837023e-07,
-      "long_answer_loss": 0.0673,
-      "loss": 0.0661,
-      "short_answer_loss": NaN,
-      "step": 1122,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0617,
-      "grad_norm": 1.4609375,
-      "learning_rate": 5.448070792661533e-07,
-      "long_answer_loss": 0.0617,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 1123,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.072,
-      "grad_norm": 1.4140625,
-      "learning_rate": 5.352762568183067e-07,
-      "long_answer_loss": 0.072,
-      "loss": 0.0719,
-      "short_answer_loss": NaN,
-      "step": 1124,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0719,
-      "grad_norm": 1.53125,
-      "learning_rate": 5.258277132816388e-07,
-      "long_answer_loss": 0.0719,
-      "loss": 0.0691,
-      "short_answer_loss": NaN,
-      "step": 1125,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0673,
-      "grad_norm": 1.4375,
-      "learning_rate": 5.164615136318163e-07,
-      "long_answer_loss": 0.0673,
-      "loss": 0.0663,
-      "short_answer_loss": NaN,
-      "step": 1126,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0696,
-      "grad_norm": 1.515625,
-      "learning_rate": 5.071777222782417e-07,
-      "long_answer_loss": 0.0696,
-      "loss": 0.0656,
-      "short_answer_loss": NaN,
-      "step": 1127,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.82,
-      "full_loss": 0.0713,
-      "grad_norm": 1.328125,
-      "learning_rate": 4.979764030636116e-07,
-      "long_answer_loss": 0.0713,
-      "loss": 0.0634,
-      "short_answer_loss": NaN,
-      "step": 1128,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0627,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.888576192634817e-07,
-      "long_answer_loss": 0.0627,
-      "loss": 0.0661,
-      "short_answer_loss": NaN,
-      "step": 1129,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0769,
-      "grad_norm": 1.515625,
-      "learning_rate": 4.798214335858267e-07,
-      "long_answer_loss": 0.0769,
-      "loss": 0.066,
-      "short_answer_loss": NaN,
-      "step": 1130,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0609,
-      "grad_norm": 1.71875,
-      "learning_rate": 4.708679081706136e-07,
-      "long_answer_loss": 0.0609,
-      "loss": 0.0662,
-      "short_answer_loss": NaN,
-      "step": 1131,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0563,
-      "grad_norm": 1.5234375,
-      "learning_rate": 4.6199710458936644e-07,
-      "long_answer_loss": 0.0563,
-      "loss": 0.0627,
-      "short_answer_loss": NaN,
-      "step": 1132,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0636,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.532090838447564e-07,
-      "long_answer_loss": 0.0636,
-      "loss": 0.0637,
-      "short_answer_loss": NaN,
-      "step": 1133,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.83,
-      "full_loss": 0.0653,
-      "grad_norm": 1.53125,
-      "learning_rate": 4.4450390637016946e-07,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 1134,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0794,
-      "grad_norm": 1.4375,
-      "learning_rate": 4.358816320292947e-07,
-      "long_answer_loss": 0.0794,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1135,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0536,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.273423201157159e-07,
-      "long_answer_loss": 0.0536,
-      "loss": 0.0651,
-      "short_answer_loss": NaN,
-      "step": 1136,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0628,
-      "grad_norm": 1.46875,
-      "learning_rate": 4.1888602935250267e-07,
-      "long_answer_loss": 0.0628,
-      "loss": 0.0664,
-      "short_answer_loss": NaN,
-      "step": 1137,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0555,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.105128178918033e-07,
-      "long_answer_loss": 0.0555,
-      "loss": 0.0693,
-      "short_answer_loss": NaN,
-      "step": 1138,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0888,
-      "grad_norm": 1.4453125,
-      "learning_rate": 4.022227433144468e-07,
-      "long_answer_loss": 0.0888,
-      "loss": 0.0665,
-      "short_answer_loss": NaN,
-      "step": 1139,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0606,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.940158626295501e-07,
-      "long_answer_loss": 0.0606,
-      "loss": 0.0662,
-      "short_answer_loss": NaN,
-      "step": 1140,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.84,
-      "full_loss": 0.0557,
-      "grad_norm": 1.4609375,
-      "learning_rate": 3.858922322741182e-07,
-      "long_answer_loss": 0.0557,
-      "loss": 0.0664,
-      "short_answer_loss": NaN,
-      "step": 1141,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0565,
-      "grad_norm": 1.390625,
-      "learning_rate": 3.778519081126641e-07,
-      "long_answer_loss": 0.0565,
-      "loss": 0.0645,
-      "short_answer_loss": NaN,
-      "step": 1142,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0817,
-      "grad_norm": 1.578125,
-      "learning_rate": 3.698949454368231e-07,
-      "long_answer_loss": 0.0817,
-      "loss": 0.0759,
-      "short_answer_loss": NaN,
-      "step": 1143,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0748,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.620213989649679e-07,
-      "long_answer_loss": 0.0748,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 1144,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0776,
-      "grad_norm": 1.5234375,
-      "learning_rate": 3.542313228418359e-07,
-      "long_answer_loss": 0.0776,
-      "loss": 0.0682,
-      "short_answer_loss": NaN,
-      "step": 1145,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0691,
-      "grad_norm": 1.375,
-      "learning_rate": 3.4652477063815833e-07,
-      "long_answer_loss": 0.0691,
-      "loss": 0.0669,
-      "short_answer_loss": NaN,
-      "step": 1146,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.85,
-      "full_loss": 0.0666,
-      "grad_norm": 1.5234375,
-      "learning_rate": 3.3890179535028544e-07,
-      "long_answer_loss": 0.0666,
-      "loss": 0.0768,
-      "short_answer_loss": NaN,
-      "step": 1147,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0666,
-      "grad_norm": 1.359375,
-      "learning_rate": 3.313624493998316e-07,
-      "long_answer_loss": 0.0666,
-      "loss": 0.0599,
-      "short_answer_loss": NaN,
-      "step": 1148,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0711,
-      "grad_norm": 1.453125,
-      "learning_rate": 3.2390678463330713e-07,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 1149,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0645,
-      "grad_norm": 1.40625,
-      "learning_rate": 3.165348523217634e-07,
-      "long_answer_loss": 0.0645,
-      "loss": 0.0642,
-      "short_answer_loss": NaN,
-      "step": 1150,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0717,
-      "grad_norm": 1.4296875,
-      "learning_rate": 3.092467031604443e-07,
-      "long_answer_loss": 0.0717,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 1151,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.077,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.0204238726842945e-07,
-      "long_answer_loss": 0.077,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1152,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.86,
-      "full_loss": 0.0727,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.9492195418829997e-07,
-      "long_answer_loss": 0.0727,
-      "loss": 0.0674,
-      "short_answer_loss": NaN,
-      "step": 1153,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0613,
-      "grad_norm": 1.40625,
-      "learning_rate": 2.878854528857888e-07,
-      "long_answer_loss": 0.0613,
-      "loss": 0.0721,
-      "short_answer_loss": NaN,
-      "step": 1154,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0614,
-      "grad_norm": 1.375,
-      "learning_rate": 2.8093293174944883e-07,
-      "long_answer_loss": 0.0614,
-      "loss": 0.0649,
-      "short_answer_loss": NaN,
-      "step": 1155,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0862,
-      "grad_norm": 1.515625,
-      "learning_rate": 2.740644385903199e-07,
-      "long_answer_loss": 0.0862,
-      "loss": 0.0664,
-      "short_answer_loss": NaN,
-      "step": 1156,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0653,
-      "grad_norm": 1.4921875,
-      "learning_rate": 2.672800206415971e-07,
-      "long_answer_loss": 0.0653,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 1157,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0598,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.605797245583075e-07,
-      "long_answer_loss": 0.0598,
-      "loss": 0.0659,
-      "short_answer_loss": NaN,
-      "step": 1158,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.87,
-      "full_loss": 0.0685,
-      "grad_norm": 1.453125,
-      "learning_rate": 2.5396359641699093e-07,
-      "long_answer_loss": 0.0685,
-      "loss": 0.0667,
-      "short_answer_loss": NaN,
-      "step": 1159,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0711,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.474316817153821e-07,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1160,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0771,
-      "grad_norm": 1.4296875,
-      "learning_rate": 2.4098402537209577e-07,
-      "long_answer_loss": 0.0771,
-      "loss": 0.0703,
-      "short_answer_loss": NaN,
-      "step": 1161,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0707,
-      "grad_norm": 1.4375,
-      "learning_rate": 2.3462067172632246e-07,
-      "long_answer_loss": 0.0707,
-      "loss": 0.065,
-      "short_answer_loss": NaN,
-      "step": 1162,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.072,
-      "grad_norm": 1.4765625,
-      "learning_rate": 2.2834166453751805e-07,
-      "long_answer_loss": 0.072,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 1163,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0784,
-      "grad_norm": 1.46875,
-      "learning_rate": 2.2214704698510503e-07,
-      "long_answer_loss": 0.0784,
-      "loss": 0.0681,
-      "short_answer_loss": NaN,
-      "step": 1164,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.88,
-      "full_loss": 0.0633,
-      "grad_norm": 1.4140625,
-      "learning_rate": 2.160368616681785e-07,
-      "long_answer_loss": 0.0633,
-      "loss": 0.062,
-      "short_answer_loss": NaN,
-      "step": 1165,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0567,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.1001115060520772e-07,
-      "long_answer_loss": 0.0567,
-      "loss": 0.0656,
-      "short_answer_loss": NaN,
-      "step": 1166,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0799,
-      "grad_norm": 1.359375,
-      "learning_rate": 2.04069955233753e-07,
-      "long_answer_loss": 0.0799,
-      "loss": 0.0628,
-      "short_answer_loss": NaN,
-      "step": 1167,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0707,
-      "grad_norm": 1.5546875,
-      "learning_rate": 1.9821331641017572e-07,
-      "long_answer_loss": 0.0707,
-      "loss": 0.07,
-      "short_answer_loss": NaN,
-      "step": 1168,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0638,
-      "grad_norm": 1.375,
-      "learning_rate": 1.9244127440936066e-07,
-      "long_answer_loss": 0.0638,
-      "loss": 0.0649,
-      "short_answer_loss": NaN,
-      "step": 1169,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0837,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.8675386892443858e-07,
-      "long_answer_loss": 0.0837,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 1170,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0755,
-      "grad_norm": 1.5390625,
-      "learning_rate": 1.8115113906650856e-07,
-      "long_answer_loss": 0.0755,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 1171,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.89,
-      "full_loss": 0.0655,
-      "grad_norm": 1.4453125,
-      "learning_rate": 1.7563312336437848e-07,
-      "long_answer_loss": 0.0655,
-      "loss": 0.0668,
-      "short_answer_loss": NaN,
-      "step": 1172,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0614,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.7019985976429174e-07,
-      "long_answer_loss": 0.0614,
-      "loss": 0.0653,
-      "short_answer_loss": NaN,
-      "step": 1173,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0657,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.6485138562966906e-07,
-      "long_answer_loss": 0.0657,
-      "loss": 0.0665,
-      "short_answer_loss": NaN,
-      "step": 1174,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0609,
-      "grad_norm": 1.421875,
-      "learning_rate": 1.5958773774085166e-07,
-      "long_answer_loss": 0.0609,
-      "loss": 0.0707,
-      "short_answer_loss": NaN,
-      "step": 1175,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0635,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.5440895229485026e-07,
-      "long_answer_loss": 0.0635,
-      "loss": 0.0637,
-      "short_answer_loss": NaN,
-      "step": 1176,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0547,
-      "grad_norm": 1.5078125,
-      "learning_rate": 1.493150649050923e-07,
-      "long_answer_loss": 0.0547,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 1177,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.9,
-      "full_loss": 0.0559,
-      "grad_norm": 1.53125,
-      "learning_rate": 1.4430611060117922e-07,
-      "long_answer_loss": 0.0559,
-      "loss": 0.0781,
-      "short_answer_loss": NaN,
-      "step": 1178,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0682,
-      "grad_norm": 1.4921875,
-      "learning_rate": 1.3938212382864497e-07,
-      "long_answer_loss": 0.0682,
-      "loss": 0.0741,
-      "short_answer_loss": NaN,
-      "step": 1179,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0589,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.345431384487214e-07,
-      "long_answer_loss": 0.0589,
-      "loss": 0.0646,
-      "short_answer_loss": NaN,
-      "step": 1180,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.073,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.2978918773810243e-07,
-      "long_answer_loss": 0.073,
-      "loss": 0.0696,
-      "short_answer_loss": NaN,
-      "step": 1181,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.06,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.251203043887164e-07,
-      "long_answer_loss": 0.06,
-      "loss": 0.0728,
-      "short_answer_loss": NaN,
-      "step": 1182,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0635,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.2053652050749846e-07,
-      "long_answer_loss": 0.0635,
-      "loss": 0.0673,
-      "short_answer_loss": NaN,
-      "step": 1183,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.91,
-      "full_loss": 0.0641,
-      "grad_norm": 1.5,
-      "learning_rate": 1.160378676161783e-07,
-      "long_answer_loss": 0.0641,
-      "loss": 0.0723,
-      "short_answer_loss": NaN,
-      "step": 1184,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0841,
-      "grad_norm": 1.453125,
-      "learning_rate": 1.1162437665105108e-07,
-      "long_answer_loss": 0.0841,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 1185,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0664,
-      "grad_norm": 1.390625,
-      "learning_rate": 1.0729607796277629e-07,
-      "long_answer_loss": 0.0664,
-      "loss": 0.0669,
-      "short_answer_loss": NaN,
-      "step": 1186,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0783,
-      "grad_norm": 1.3515625,
-      "learning_rate": 1.0305300131616125e-07,
-      "long_answer_loss": 0.0783,
-      "loss": 0.0656,
-      "short_answer_loss": NaN,
-      "step": 1187,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0659,
-      "grad_norm": 1.3671875,
-      "learning_rate": 9.889517588995839e-08,
-      "long_answer_loss": 0.0659,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 1188,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0664,
-      "grad_norm": 1.5234375,
-      "learning_rate": 9.482263027666833e-08,
-      "long_answer_loss": 0.0664,
-      "loss": 0.0708,
-      "short_answer_loss": NaN,
-      "step": 1189,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.92,
-      "full_loss": 0.0586,
-      "grad_norm": 1.3359375,
-      "learning_rate": 9.083539248233852e-08,
-      "long_answer_loss": 0.0586,
-      "loss": 0.0622,
-      "short_answer_loss": NaN,
-      "step": 1190,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0623,
-      "grad_norm": 1.4453125,
-      "learning_rate": 8.693348992637046e-08,
-      "long_answer_loss": 0.0623,
-      "loss": 0.0672,
-      "short_answer_loss": NaN,
-      "step": 1191,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0905,
-      "grad_norm": 1.4921875,
-      "learning_rate": 8.311694944133502e-08,
-      "long_answer_loss": 0.0905,
-      "loss": 0.0699,
-      "short_answer_loss": NaN,
-      "step": 1192,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0673,
-      "grad_norm": 1.484375,
-      "learning_rate": 7.938579727278517e-08,
-      "long_answer_loss": 0.0673,
-      "loss": 0.068,
-      "short_answer_loss": NaN,
-      "step": 1193,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0676,
-      "grad_norm": 1.5078125,
-      "learning_rate": 7.574005907907966e-08,
-      "long_answer_loss": 0.0676,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 1194,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0656,
-      "grad_norm": 1.3984375,
-      "learning_rate": 7.217975993119713e-08,
-      "long_answer_loss": 0.0656,
-      "loss": 0.0641,
-      "short_answer_loss": NaN,
-      "step": 1195,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.93,
-      "full_loss": 0.0809,
-      "grad_norm": 1.390625,
-      "learning_rate": 6.87049243125723e-08,
-      "long_answer_loss": 0.0809,
-      "loss": 0.0661,
-      "short_answer_loss": NaN,
-      "step": 1196,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0564,
-      "grad_norm": 1.484375,
-      "learning_rate": 6.531557611892669e-08,
-      "long_answer_loss": 0.0564,
-      "loss": 0.0698,
-      "short_answer_loss": NaN,
-      "step": 1197,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0766,
-      "grad_norm": 1.4140625,
-      "learning_rate": 6.201173865810207e-08,
-      "long_answer_loss": 0.0766,
-      "loss": 0.0637,
-      "short_answer_loss": NaN,
-      "step": 1198,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0868,
-      "grad_norm": 1.8046875,
-      "learning_rate": 5.879343464989806e-08,
-      "long_answer_loss": 0.0868,
-      "loss": 0.0755,
-      "short_answer_loss": NaN,
-      "step": 1199,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0726,
-      "grad_norm": 1.4453125,
-      "learning_rate": 5.566068622592235e-08,
-      "long_answer_loss": 0.0726,
-      "loss": 0.0702,
-      "short_answer_loss": NaN,
-      "step": 1200,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0771,
-      "grad_norm": 1.5,
-      "learning_rate": 5.261351492943101e-08,
-      "long_answer_loss": 0.0771,
-      "loss": 0.0651,
-      "short_answer_loss": NaN,
-      "step": 1201,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.94,
-      "full_loss": 0.0595,
-      "grad_norm": 1.53125,
-      "learning_rate": 4.965194171518833e-08,
-      "long_answer_loss": 0.0595,
-      "loss": 0.0685,
-      "short_answer_loss": NaN,
-      "step": 1202,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0644,
-      "grad_norm": 1.4609375,
-      "learning_rate": 4.677598694931285e-08,
-      "long_answer_loss": 0.0644,
-      "loss": 0.0731,
-      "short_answer_loss": NaN,
-      "step": 1203,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0676,
-      "grad_norm": 1.453125,
-      "learning_rate": 4.3985670409148196e-08,
-      "long_answer_loss": 0.0676,
-      "loss": 0.0671,
-      "short_answer_loss": NaN,
-      "step": 1204,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0634,
-      "grad_norm": 1.390625,
-      "learning_rate": 4.128101128312023e-08,
-      "long_answer_loss": 0.0634,
-      "loss": 0.071,
-      "short_answer_loss": NaN,
-      "step": 1205,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0868,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.866202817060377e-08,
-      "long_answer_loss": 0.0868,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 1206,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0788,
-      "grad_norm": 1.4453125,
-      "learning_rate": 3.612873908180048e-08,
-      "long_answer_loss": 0.0788,
-      "loss": 0.0697,
-      "short_answer_loss": NaN,
-      "step": 1207,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.073,
-      "grad_norm": 1.609375,
-      "learning_rate": 3.3681161437612575e-08,
-      "long_answer_loss": 0.073,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1208,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.95,
-      "full_loss": 0.0587,
-      "grad_norm": 1.421875,
-      "learning_rate": 3.131931206951933e-08,
-      "long_answer_loss": 0.0587,
-      "loss": 0.064,
-      "short_answer_loss": NaN,
-      "step": 1209,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.053,
-      "grad_norm": 1.5390625,
-      "learning_rate": 2.9043207219468795e-08,
-      "long_answer_loss": 0.053,
-      "loss": 0.0679,
-      "short_answer_loss": NaN,
-      "step": 1210,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0482,
-      "grad_norm": 1.5625,
-      "learning_rate": 2.6852862539757106e-08,
-      "long_answer_loss": 0.0482,
-      "loss": 0.0711,
-      "short_answer_loss": NaN,
-      "step": 1211,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0736,
-      "grad_norm": 1.40625,
-      "learning_rate": 2.4748293092931308e-08,
-      "long_answer_loss": 0.0736,
-      "loss": 0.067,
-      "short_answer_loss": NaN,
-      "step": 1212,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0611,
-      "grad_norm": 1.46875,
-      "learning_rate": 2.2729513351672783e-08,
-      "long_answer_loss": 0.0611,
-      "loss": 0.0718,
-      "short_answer_loss": NaN,
-      "step": 1213,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0569,
-      "grad_norm": 1.4609375,
-      "learning_rate": 2.0796537198712608e-08,
-      "long_answer_loss": 0.0569,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1214,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.96,
-      "full_loss": 0.0782,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.894937792672191e-08,
-      "long_answer_loss": 0.0782,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 1215,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0704,
-      "grad_norm": 1.515625,
-      "learning_rate": 1.7188048238232778e-08,
-      "long_answer_loss": 0.0704,
-      "loss": 0.0732,
-      "short_answer_loss": NaN,
-      "step": 1216,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0611,
-      "grad_norm": 1.40625,
-      "learning_rate": 1.5512560245541097e-08,
-      "long_answer_loss": 0.0611,
-      "loss": 0.0676,
-      "short_answer_loss": NaN,
-      "step": 1217,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.062,
-      "grad_norm": 1.5,
-      "learning_rate": 1.3922925470627458e-08,
-      "long_answer_loss": 0.062,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1218,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0516,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.2419154845079439e-08,
-      "long_answer_loss": 0.0516,
-      "loss": 0.0623,
-      "short_answer_loss": NaN,
-      "step": 1219,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0778,
-      "grad_norm": 1.4765625,
-      "learning_rate": 1.1001258710015283e-08,
-      "long_answer_loss": 0.0778,
-      "loss": 0.0637,
-      "short_answer_loss": NaN,
-      "step": 1220,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.97,
-      "full_loss": 0.0571,
-      "grad_norm": 1.484375,
-      "learning_rate": 9.669246816010335e-09,
-      "long_answer_loss": 0.0571,
-      "loss": 0.0676,
-      "short_answer_loss": NaN,
-      "step": 1221,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0721,
-      "grad_norm": 1.4140625,
-      "learning_rate": 8.423128323033213e-09,
-      "long_answer_loss": 0.0721,
-      "loss": 0.0687,
-      "short_answer_loss": NaN,
-      "step": 1222,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0665,
-      "grad_norm": 1.515625,
-      "learning_rate": 7.262911800379191e-09,
-      "long_answer_loss": 0.0665,
-      "loss": 0.073,
-      "short_answer_loss": NaN,
-      "step": 1223,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0723,
-      "grad_norm": 1.5078125,
-      "learning_rate": 6.188605226618849e-09,
-      "long_answer_loss": 0.0723,
-      "loss": 0.0722,
-      "short_answer_loss": NaN,
-      "step": 1224,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0667,
-      "grad_norm": 1.390625,
-      "learning_rate": 5.200215989531465e-09,
-      "long_answer_loss": 0.0667,
-      "loss": 0.0653,
-      "short_answer_loss": NaN,
-      "step": 1225,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.061,
-      "grad_norm": 1.421875,
-      "learning_rate": 4.297750886064766e-09,
-      "long_answer_loss": 0.061,
-      "loss": 0.0637,
-      "short_answer_loss": NaN,
-      "step": 1226,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.98,
-      "full_loss": 0.0719,
-      "grad_norm": 1.484375,
-      "learning_rate": 3.481216122284969e-09,
-      "long_answer_loss": 0.0719,
-      "loss": 0.0714,
-      "short_answer_loss": NaN,
-      "step": 1227,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0724,
-      "grad_norm": 1.421875,
-      "learning_rate": 2.7506173133282075e-09,
-      "long_answer_loss": 0.0724,
-      "loss": 0.0656,
-      "short_answer_loss": NaN,
-      "step": 1228,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0666,
-      "grad_norm": 1.515625,
-      "learning_rate": 2.105959483371389e-09,
-      "long_answer_loss": 0.0666,
-      "loss": 0.0666,
-      "short_answer_loss": NaN,
-      "step": 1229,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0949,
-      "grad_norm": 1.4375,
-      "learning_rate": 1.547247065593338e-09,
-      "long_answer_loss": 0.0949,
-      "loss": 0.0695,
-      "short_answer_loss": NaN,
-      "step": 1230,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0807,
-      "grad_norm": 1.4296875,
-      "learning_rate": 1.0744839021428755e-09,
-      "long_answer_loss": 0.0807,
-      "loss": 0.0709,
-      "short_answer_loss": NaN,
-      "step": 1231,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0711,
-      "grad_norm": 1.5,
-      "learning_rate": 6.876732441110645e-10,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0677,
-      "short_answer_loss": NaN,
-      "step": 1232,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 1.99,
-      "full_loss": 0.0797,
-      "grad_norm": 1.46875,
-      "learning_rate": 3.868177515173321e-10,
-      "long_answer_loss": 0.0797,
-      "loss": 0.0694,
-      "short_answer_loss": NaN,
-      "step": 1233,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "full_loss": 0.0711,
-      "grad_norm": 1.4140625,
-      "learning_rate": 1.7191949328032587e-10,
-      "long_answer_loss": 0.0711,
-      "loss": 0.0705,
-      "short_answer_loss": NaN,
-      "step": 1234,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "full_loss": 0.0733,
-      "grad_norm": 1.375,
-      "learning_rate": 4.297994721097487e-11,
-      "long_answer_loss": 0.0733,
-      "loss": 0.0645,
-      "short_answer_loss": NaN,
-      "step": 1235,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "full_loss": 0.0661,
-      "grad_norm": 1.65625,
-      "learning_rate": 0.0,
-      "long_answer_loss": 0.0661,
-      "loss": 0.0717,
-      "short_answer_loss": NaN,
-      "step": 1236,
-      "template_loss": 0.0
-    },
-    {
-      "epoch": 2.0,
-      "step": 1236,
-      "total_flos": 9.29924110325121e+17,
-      "train_loss": 0.12603917728482616,
-      "train_runtime": 5247.2889,
-      "train_samples_per_second": 30.175,
-      "train_steps_per_second": 0.236
-    }
-  ],
-  "logging_steps": 1.0,
-  "max_steps": 1236,
-  "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
-  "save_steps": 1.0,
-  "total_flos": 9.29924110325121e+17,
-  "train_batch_size": 1,
-  "trial_name": null,
-  "trial_params": null
-}