diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6640 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.976631448884397,
+  "global_step": 1090500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4e-05,
+      "loss": 9.6608,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8e-05,
+      "loss": 8.6223,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00012,
+      "loss": 8.3175,
+      "step": 300
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00016,
+      "loss": 7.9745,
+      "step": 400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0002,
+      "loss": 7.6776,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00024,
+      "loss": 7.4451,
+      "step": 600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00028,
+      "loss": 7.2587,
+      "step": 700
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00032,
+      "loss": 7.0977,
+      "step": 800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00036,
+      "loss": 6.9377,
+      "step": 900
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0004,
+      "loss": 6.8182,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999999998815762,
+      "loss": 6.6945,
+      "step": 1100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999999995263047,
+      "loss": 6.5851,
+      "step": 1200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999999893418564,
+      "loss": 6.476,
+      "step": 1300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999999981052189,
+      "loss": 6.3753,
+      "step": 1400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999999703940455,
+      "loss": 6.2997,
+      "step": 1500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999933291862616,
+      "loss": 5.9559,
+      "step": 2000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999994734068435,
+      "loss": 6.1649,
+      "step": 3000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999988151660478,
+      "loss": 5.8819,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000399997893630147,
+      "loss": 5.8437,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003999967088003543,
+      "loss": 5.857,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999526067822954,
+      "loss": 5.7574,
+      "step": 7000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00039999354926567907,
+      "loss": 5.6647,
+      "step": 8000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039999157456495604,
+      "loss": 5.598,
+      "step": 9000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039998933657865997,
+      "loss": 5.528,
+      "step": 10000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00039998683530973725,
+      "loss": 5.4848,
+      "step": 11000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999840707614807,
+      "loss": 5.4314,
+      "step": 12000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0003999810429375299,
+      "loss": 5.3931,
+      "step": 13000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039997775526446917,
+      "loss": 5.3531,
+      "step": 14000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00039997420116469963,
+      "loss": 5.3126,
+      "step": 15000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003999703877510894,
+      "loss": 5.2782,
+      "step": 16000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000399966307394198,
+      "loss": 5.2575,
+      "step": 17000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003999619682600994,
+      "loss": 5.2297,
+      "step": 18000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003999573616675516,
+      "loss": 5.1976,
+      "step": 19000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039995249683579117,
+      "loss": 5.182,
+      "step": 20000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039994736403182074,
+      "loss": 5.1576,
+      "step": 21000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039994197352799087,
+      "loss": 5.1435,
+      "step": 22000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999363145395998,
+      "loss": 5.1286,
+      "step": 23000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999303983920581,
+      "loss": 5.1065,
+      "step": 24000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003999242132490164,
+      "loss": 5.0946,
+      "step": 25000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003999177714888857,
+      "loss": 5.0748,
+      "step": 26000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039991106022373136,
+      "loss": 5.0674,
+      "step": 27000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003999040928848998,
+      "loss": 5.0493,
+      "step": 28000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998968555329385,
+      "loss": 5.039,
+      "step": 29000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998893626520587,
+      "loss": 5.0348,
+      "step": 30000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003998815992513638,
+      "loss": 5.0201,
+      "step": 31000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039987357270987667,
+      "loss": 5.0161,
+      "step": 32000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003998652914592657,
+      "loss": 4.9991,
+      "step": 33000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039985673893135445,
+      "loss": 4.9971,
+      "step": 34000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003998479232953792,
+      "loss": 4.9871,
+      "step": 35000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003998388445629455,
+      "loss": 4.9771,
+      "step": 36000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003998295122192289,
+      "loss": 4.9726,
+      "step": 37000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003998199075931465,
+      "loss": 4.9669,
+      "step": 38000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003998100499065675,
+      "loss": 4.9583,
+      "step": 39000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039979992969921984,
+      "loss": 4.9556,
+      "step": 40000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003997895364597799,
+      "loss": 4.942,
+      "step": 41000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997788909883795,
+      "loss": 4.9406,
+      "step": 42000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039976797198678043,
+      "loss": 4.9323,
+      "step": 43000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039975680130732954,
+      "loss": 4.9277,
+      "step": 44000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997453566010126,
+      "loss": 4.923,
+      "step": 45000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997336489332646,
+      "loss": 4.9197,
+      "step": 46000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003997216904214485,
+      "loss": 4.9051,
+      "step": 47000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039970945714034553,
+      "loss": 4.9077,
+      "step": 48000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003996969609450725,
+      "loss": 4.9002,
+      "step": 49000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003996842147424852,
+      "loss": 4.9013,
+      "step": 50000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039967119303144363,
+      "loss": 4.8946,
+      "step": 51000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039965792187247553,
+      "loss": 4.8882,
+      "step": 52000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039964437471416833,
+      "loss": 4.8894,
+      "step": 53000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003996305786686345,
+      "loss": 4.8764,
+      "step": 54000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003996165061343288,
+      "loss": 4.8782,
+      "step": 55000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003996021852748057,
+      "loss": 4.8759,
+      "step": 56000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039958758743853225,
+      "loss": 4.8727,
+      "step": 57000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003995727418403572,
+      "loss": 4.8669,
+      "step": 58000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003995576187789104,
+      "loss": 4.8694,
+      "step": 59000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039954224852018107,
+      "loss": 4.8688,
+      "step": 60000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003995266003131184,
+      "loss": 4.86,
+      "step": 61000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039951070547469266,
+      "loss": 4.8559,
+      "step": 62000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039949453220433417,
+      "loss": 4.8543,
+      "step": 63000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039947811286982935,
+      "loss": 4.8515,
+      "step": 64000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003994614146212571,
+      "loss": 4.8498,
+      "step": 65000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039944447087704996,
+      "loss": 4.8443,
+      "step": 66000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003994272477381079,
+      "loss": 4.8399,
+      "step": 67000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003994097796733338,
+      "loss": 4.8381,
+      "step": 68000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039939203173462723,
+      "loss": 4.8381,
+      "step": 69000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039937403944117984,
+      "loss": 4.8354,
+      "step": 70000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039935576679607466,
+      "loss": 4.834,
+      "step": 71000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003993372503686054,
+      "loss": 4.8337,
+      "step": 72000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003993184531132279,
+      "loss": 4.8304,
+      "step": 73000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039929939345843064,
+      "loss": 4.8254,
+      "step": 74000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000399280090882382,
+      "loss": 4.8248,
+      "step": 75000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003992605067667017,
+      "loss": 4.8248,
+      "step": 76000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003992406803053476,
+      "loss": 4.8246,
+      "step": 77000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039922057183181,
+      "loss": 4.8173,
+      "step": 78000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00039920024209092803,
+      "loss": 4.8128,
+      "step": 79000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039917960962754717,
+      "loss": 4.818,
+      "step": 80000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039915871494753167,
+      "loss": 4.8107,
+      "step": 81000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039913755807838893,
+      "loss": 4.8121,
+      "step": 82000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00039911613904797174,
+      "loss": 4.8116,
+      "step": 83000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003990944796965674,
+      "loss": 4.8057,
+      "step": 84000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003990725366906298,
+      "loss": 4.8055,
+      "step": 85000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003990503762807127,
+      "loss": 4.8028,
+      "step": 86000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039902790967672147,
+      "loss": 4.7969,
+      "step": 87000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003990052039152944,
+      "loss": 4.8025,
+      "step": 88000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039898221356934855,
+      "loss": 4.8017,
+      "step": 89000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00039895896126663653,
+      "loss": 4.7986,
+      "step": 90000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003989354470377698,
+      "loss": 4.7991,
+      "step": 91000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00039891169482063473,
+      "loss": 4.7965,
+      "step": 92000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00039888765709451975,
+      "loss": 4.792,
+      "step": 93000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00039886338196645364,
+      "loss": 4.7862,
+      "step": 94000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00039883882086954475,
+      "loss": 4.7916,
+      "step": 95000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000398814022958251,
+      "loss": 4.7883,
+      "step": 96000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00039878893861975594,
+      "loss": 4.7908,
+      "step": 97000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00039876359257893807,
+      "loss": 4.7877,
+      "step": 98000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003987379848691651,
+      "loss": 4.7873,
+      "step": 99000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00039871214152416957,
+      "loss": 4.7876,
+      "step": 100000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00039868601083955114,
+      "loss": 4.7883,
+      "step": 101000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00039865964511100514,
+      "loss": 4.7893,
+      "step": 102000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003986329915890061,
+      "loss": 4.7789,
+      "step": 103000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00039860610361561096,
+      "loss": 4.7815,
+      "step": 104000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003985789273964466,
+      "loss": 4.7738,
+      "step": 105000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003985515173196509,
+      "loss": 4.7753,
+      "step": 106000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00039852381854628627,
+      "loss": 4.7724,
+      "step": 107000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00039849588651028544,
+      "loss": 4.7726,
+      "step": 108000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003984676653284346,
+      "loss": 4.7685,
+      "step": 109000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003984392114801697,
+      "loss": 4.7715,
+      "step": 110000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003984104680382948,
+      "loss": 4.7713,
+      "step": 111000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00039838149252745204,
+      "loss": 4.7698,
+      "step": 112000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003983522269767629,
+      "loss": 4.7753,
+      "step": 113000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00039832272995577275,
+      "loss": 4.7652,
+      "step": 114000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003982929424502255,
+      "loss": 4.7664,
+      "step": 115000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00039826292407426207,
+      "loss": 4.7713,
+      "step": 116000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00039823264521022384,
+      "loss": 4.7628,
+      "step": 117000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003982020751975389,
+      "loss": 4.7682,
+      "step": 118000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00039817124425512714,
+      "loss": 4.7644,
+      "step": 119000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003981401524235768,
+      "loss": 4.758,
+      "step": 120000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00039810883122677967,
+      "loss": 4.7622,
+      "step": 121000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003980772180008777,
+      "loss": 4.762,
+      "step": 122000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003980453760138509,
+      "loss": 4.7571,
+      "step": 123000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00039801324155990393,
+      "loss": 4.7619,
+      "step": 124000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00039798091147522796,
+      "loss": 4.7618,
+      "step": 125000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003979482562229017,
+      "loss": 4.762,
+      "step": 126000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000397915340374997,
+      "loss": 4.7562,
+      "step": 127000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00039788216397484706,
+      "loss": 4.7528,
+      "step": 128000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00039784876063314606,
+      "loss": 4.7567,
+      "step": 129000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00039781506352031947,
+      "loss": 4.7554,
+      "step": 130000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00039778114007485855,
+      "loss": 4.7494,
+      "step": 131000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00039774692242662465,
+      "loss": 4.7591,
+      "step": 132000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00039771244444786484,
+      "loss": 4.7605,
+      "step": 133000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003976777061839689,
+      "loss": 4.7469,
+      "step": 134000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00039764274280914674,
+      "loss": 4.7506,
+      "step": 135000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00039760748437268835,
+      "loss": 4.7506,
+      "step": 136000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003975720014377832,
+      "loss": 4.7509,
+      "step": 137000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00039753622301424524,
+      "loss": 4.7488,
+      "step": 138000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00039750022070592105,
+      "loss": 4.7544,
+      "step": 139000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003974639224835218,
+      "loss": 4.7502,
+      "step": 140000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003974274009911748,
+      "loss": 4.7433,
+      "step": 141000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00039739058316086716,
+      "loss": 4.7466,
+      "step": 142000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003973535798838411,
+      "loss": 4.7469,
+      "step": 143000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003973162428990996,
+      "loss": 4.7414,
+      "step": 144000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00039727864615081464,
+      "loss": 4.7418,
+      "step": 145000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003972407896884818,
+      "loss": 4.7484,
+      "step": 146000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00039720271180775053,
+      "loss": 4.7454,
+      "step": 147000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003971643363267646,
+      "loss": 4.744,
+      "step": 148000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00039712577881131754,
+      "loss": 4.7369,
+      "step": 149000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00039708688477304655,
+      "loss": 4.7375,
+      "step": 150000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003970477312731783,
+      "loss": 4.7414,
+      "step": 151000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003970083183632576,
+      "loss": 4.7389,
+      "step": 152000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003969686858969712,
+      "loss": 4.7378,
+      "step": 153000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003969287545822263,
+      "loss": 4.7372,
+      "step": 154000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00039688860433410763,
+      "loss": 4.7393,
+      "step": 155000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00039684815482460387,
+      "loss": 4.7315,
+      "step": 156000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00039680748700586993,
+      "loss": 4.7371,
+      "step": 157000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00039676651951439873,
+      "loss": 4.7353,
+      "step": 158000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003967253343389894,
+      "loss": 4.7315,
+      "step": 159000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00039668384908106706,
+      "loss": 4.7358,
+      "step": 160000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00039664210489213713,
+      "loss": 4.7339,
+      "step": 161000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00039660010182715526,
+      "loss": 4.737,
+      "step": 162000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000396557882332566,
+      "loss": 4.7261,
+      "step": 163000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003965153619404471,
+      "loss": 4.7342,
+      "step": 164000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003964726257474391,
+      "loss": 4.7293,
+      "step": 165000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003964295882518688,
+      "loss": 4.7301,
+      "step": 166000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00039638633558526285,
+      "loss": 4.7316,
+      "step": 167000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00039634278121264703,
+      "loss": 4.7295,
+      "step": 168000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003962990122999811,
+      "loss": 4.7332,
+      "step": 169000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003962549412794449,
+      "loss": 4.732,
+      "step": 170000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003962106563509727,
+      "loss": 4.7321,
+      "step": 171000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00039616606891435896,
+      "loss": 4.7276,
+      "step": 172000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00039612122322838677,
+      "loss": 4.7245,
+      "step": 173000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003960761645849172,
+      "loss": 4.7286,
+      "step": 174000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003960308028357847,
+      "loss": 4.7239,
+      "step": 175000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0003840903997775841,
+      "loss": 4.6145,
+      "step": 176000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0003839104648613638,
+      "loss": 4.5905,
+      "step": 177000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0003837297421617577,
+      "loss": 4.5891,
+      "step": 178000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.000383548053178735,
+      "loss": 4.5817,
+      "step": 179000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0003833652155473882,
+      "loss": 4.5765,
+      "step": 180000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00038318141161813824,
+      "loss": 4.574,
+      "step": 181000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0003829966423595951,
+      "loss": 4.5725,
+      "step": 182000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00038281109496044006,
+      "loss": 4.5666,
+      "step": 183000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00038262439893236937,
+      "loss": 4.5631,
+      "step": 184000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00038243692864915963,
+      "loss": 4.5591,
+      "step": 185000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0003822483097830243,
+      "loss": 4.5552,
+      "step": 186000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00038205873050485524,
+      "loss": 4.5543,
+      "step": 187000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0003818683828312813,
+      "loss": 4.5512,
+      "step": 188000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00038167688668914063,
+      "loss": 4.5484,
+      "step": 189000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0003814844331462512,
+      "loss": 4.5501,
+      "step": 190000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0003812912171041104,
+      "loss": 4.5431,
+      "step": 191000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0003810968527621949,
+      "loss": 4.5418,
+      "step": 192000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00038090153407619305,
+      "loss": 4.5379,
+      "step": 193000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00038070526207539536,
+      "loss": 4.538,
+      "step": 194000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0003805082354937156,
+      "loss": 4.5377,
+      "step": 195000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0003803100609220069,
+      "loss": 4.5354,
+      "step": 196000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0003801111357514916,
+      "loss": 4.5321,
+      "step": 197000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000379911062782051,
+      "loss": 4.5327,
+      "step": 198000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00037971004171739956,
+      "loss": 4.5342,
+      "step": 199000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00037950827605766894,
+      "loss": 4.527,
+      "step": 200000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00037930536293104657,
+      "loss": 4.5297,
+      "step": 201000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00037910170923078203,
+      "loss": 4.5252,
+      "step": 202000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00037889690831515295,
+      "loss": 4.5228,
+      "step": 203000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0003786911646487036,
+      "loss": 4.5211,
+      "step": 204000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00037848447931566176,
+      "loss": 4.521,
+      "step": 205000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00037827685340521773,
+      "loss": 4.5257,
+      "step": 206000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003780684970458185,
+      "loss": 4.5204,
+      "step": 207000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003778589942057952,
+      "loss": 4.5209,
+      "step": 208000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003776487649924752,
+      "loss": 4.5167,
+      "step": 209000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003774373896346034,
+      "loss": 4.5142,
+      "step": 210000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00037722507921728195,
+      "loss": 4.5166,
+      "step": 211000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003770122622793867,
+      "loss": 4.5127,
+      "step": 212000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00037679808696909655,
+      "loss": 4.5163,
+      "step": 213000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00037658297996835357,
+      "loss": 4.513,
+      "step": 214000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0003763669424107285,
+      "loss": 4.5078,
+      "step": 215000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00037614997543469595,
+      "loss": 4.5114,
+      "step": 216000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00037593208018362834,
+      "loss": 4.5097,
+      "step": 217000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00037571369637505247,
+      "loss": 4.5072,
+      "step": 218000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00037549394987438647,
+      "loss": 4.5084,
+      "step": 219000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00037527327855580843,
+      "loss": 4.5071,
+      "step": 220000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003750519056381631,
+      "loss": 4.5061,
+      "step": 221000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00037482938909921175,
+      "loss": 4.5075,
+      "step": 222000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003746059512444505,
+      "loss": 4.5079,
+      "step": 223000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0003743815932513518,
+      "loss": 4.5071,
+      "step": 224000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00037415631630223755,
+      "loss": 4.5033,
+      "step": 225000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00037393012158427186,
+      "loss": 4.505,
+      "step": 226000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00037370323785818266,
+      "loss": 4.5032,
+      "step": 227000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00037347521209812743,
+      "loss": 4.5017,
+      "step": 228000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0003732465015546745,
+      "loss": 4.502,
+      "step": 229000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00037301664955431804,
+      "loss": 4.4998,
+      "step": 230000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0003727858857909254,
+      "loss": 4.4994,
+      "step": 231000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0003725544436092979,
+      "loss": 4.4985,
+      "step": 232000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.000372321860881582,
+      "loss": 4.499,
+      "step": 233000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00037208837005222694,
+      "loss": 4.4919,
+      "step": 234000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0003718542072019544,
+      "loss": 4.4965,
+      "step": 235000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00037161890477046666,
+      "loss": 4.4972,
+      "step": 236000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00037138293459993847,
+      "loss": 4.4988,
+      "step": 237000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039974046056824423,
+      "loss": 5.0173,
+      "step": 238000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997382653105697,
+      "loss": 5.1254,
+      "step": 239000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997360630230883,
+      "loss": 5.137,
+      "step": 240000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997338515152591,
+      "loss": 5.1396,
+      "step": 241000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003997316285596137,
+      "loss": 5.1539,
+      "step": 242000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039972939860216607,
+      "loss": 5.1836,
+      "step": 243000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039972715717864,
+      "loss": 5.1907,
+      "step": 244000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039972490651670964,
+      "loss": 5.2177,
+      "step": 245000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039972264888099373,
+      "loss": 5.2218,
+      "step": 246000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003955398710520662,
+      "loss": 4.9553,
+      "step": 247000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000395503657852559,
+      "loss": 4.8679,
+      "step": 248000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00039546729990487664,
+      "loss": 4.8395,
+      "step": 249000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003954307972359379,
+      "loss": 4.8217,
+      "step": 250000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000395394149872769,
+      "loss": 4.8152,
+      "step": 251000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003953573947067854,
+      "loss": 4.8026,
+      "step": 252000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003953204581812889,
+      "loss": 4.8017,
+      "step": 253000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003952834141966186,
+      "loss": 4.7977,
+      "step": 254000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00039524618861807426,
+      "loss": 4.7963,
+      "step": 255000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003952088184819814,
+      "loss": 4.79,
+      "step": 256000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003951713414028577,
+      "loss": 4.7877,
+      "step": 257000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003951336823792677,
+      "loss": 4.7854,
+      "step": 258000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003950959167570807,
+      "loss": 4.7945,
+      "step": 259000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00039505796895741114,
+      "loss": 4.7845,
+      "step": 260000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00039501991490389356,
+      "loss": 4.7821,
+      "step": 261000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000394981678440416,
+      "loss": 4.7798,
+      "step": 262000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00039494333606815397,
+      "loss": 4.7892,
+      "step": 263000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00039490481105399416,
+      "loss": 4.7885,
+      "step": 264000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000394866141735037,
+      "loss": 4.7838,
+      "step": 265000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0003948273670255641,
+      "loss": 4.7812,
+      "step": 266000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00039478840932724265,
+      "loss": 4.7749,
+      "step": 267000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00039474934658425046,
+      "loss": 4.7823,
+      "step": 268000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00039471010062182423,
+      "loss": 4.7809,
+      "step": 269000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00039467074996088307,
+      "loss": 4.785,
+      "step": 270000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0003946312158504645,
+      "loss": 4.7753,
+      "step": 271000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00039459157738799654,
+      "loss": 4.775,
+      "step": 272000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0003945517552465506,
+      "loss": 4.7755,
+      "step": 273000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0003945118290998296,
+      "loss": 4.7849,
+      "step": 274000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00039447175922715307,
+      "loss": 4.7806,
+      "step": 275000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00039443150533232405,
+      "loss": 4.7791,
+      "step": 276000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00039439110748312647,
+      "loss": 4.7798,
+      "step": 277000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00039435056570947044,
+      "loss": 4.7794,
+      "step": 278000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0003943099207989059,
+      "loss": 4.7821,
+      "step": 279000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0003942690914103384,
+      "loss": 4.7815,
+      "step": 280000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00039422811818765134,
+      "loss": 4.7713,
+      "step": 281000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00039418704235002724,
+      "loss": 4.7707,
+      "step": 282000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00039414582302643454,
+      "loss": 4.7764,
+      "step": 283000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0003941044187712859,
+      "loss": 4.7864,
+      "step": 284000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00039406287080393925,
+      "loss": 4.774,
+      "step": 285000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0003940211791551559,
+      "loss": 4.7698,
+      "step": 286000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00039397938576284634,
+      "loss": 4.7754,
+      "step": 287000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00039393740698750394,
+      "loss": 4.7764,
+      "step": 288000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0003938952846236165,
+      "loss": 4.7764,
+      "step": 289000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00039385301870237103,
+      "loss": 4.7747,
+      "step": 290000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00039381065173618853,
+      "loss": 4.7784,
+      "step": 291000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00039376809893769117,
+      "loss": 4.7792,
+      "step": 292000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00039372544544391313,
+      "loss": 4.7726,
+      "step": 293000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0003936826488052433,
+      "loss": 4.7736,
+      "step": 294000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0003936396659988803,
+      "loss": 4.7759,
+      "step": 295000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00039359653982441555,
+      "loss": 4.7719,
+      "step": 296000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00039355327031377916,
+      "loss": 4.7775,
+      "step": 297000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0003935099009833917,
+      "loss": 4.7814,
+      "step": 298000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00039346634503988233,
+      "loss": 4.7722,
+      "step": 299000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0003934226458565957,
+      "loss": 4.7745,
+      "step": 300000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000393378847379798,
+      "loss": 4.7748,
+      "step": 301000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00039333486195728426,
+      "loss": 4.7774,
+      "step": 302000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00039329077759239523,
+      "loss": 4.7777,
+      "step": 303000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.000393246506060789,
+      "loss": 4.7707,
+      "step": 304000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0003932021359382358,
+      "loss": 4.7792,
+      "step": 305000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.000393157578428518,
+      "loss": 4.7711,
+      "step": 306000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0003931128779076294,
+      "loss": 4.7712,
+      "step": 307000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0003930680793235711,
+      "loss": 4.7732,
+      "step": 308000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00039302309302266194,
+      "loss": 4.7753,
+      "step": 309000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00039297800901073876,
+      "loss": 4.7747,
+      "step": 310000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003929327370629047,
+      "loss": 4.7756,
+      "step": 311000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003928873222703692,
+      "loss": 4.7733,
+      "step": 312000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003928418102956833,
+      "loss": 4.7714,
+      "step": 313000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003927961100574846,
+      "loss": 4.773,
+      "step": 314000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003927503129900122,
+      "loss": 4.7742,
+      "step": 315000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003927043274413583,
+      "loss": 4.7757,
+      "step": 316000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0003926582454165936,
+      "loss": 4.7738,
+      "step": 317000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00039261202103549754,
+      "loss": 4.7675,
+      "step": 318000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0003925656078478171,
+      "loss": 4.7782,
+      "step": 319000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00039251905208725256,
+      "loss": 4.7703,
+      "step": 320000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00039247235378827314,
+      "loss": 4.7726,
+      "step": 321000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0003924255598974257,
+      "loss": 4.7679,
+      "step": 322000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00039237857676789823,
+      "loss": 4.767,
+      "step": 323000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0003923314984006603,
+      "loss": 4.7621,
+      "step": 324000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0003922842305795883,
+      "loss": 4.7717,
+      "step": 325000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00039223686787524505,
+      "loss": 4.7682,
+      "step": 326000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0003921893155024742,
+      "loss": 4.783,
+      "step": 327000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0003921416686011523,
+      "loss": 4.7705,
+      "step": 328000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0003920938318173703,
+      "loss": 4.7678,
+      "step": 329000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0003920459008600368,
+      "loss": 4.7697,
+      "step": 330000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0003919977798067727,
+      "loss": 4.7749,
+      "step": 331000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00039194956493523547,
+      "loss": 4.7797,
+      "step": 332000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00039190115975485935,
+      "loss": 4.7678,
+      "step": 333000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0003918526124935473,
+      "loss": 4.7674,
+      "step": 334000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0003918039719474887,
+      "loss": 4.7711,
+      "step": 335000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0003917551407742319,
+      "loss": 4.7686,
+      "step": 336000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00039170621667219887,
+      "loss": 4.7708,
+      "step": 337000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00039165710173146836,
+      "loss": 4.7681,
+      "step": 338000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0003916078942182069,
+      "loss": 4.7697,
+      "step": 339000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0003915584956553133,
+      "loss": 4.7665,
+      "step": 340000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00039150900487640804,
+      "loss": 4.7757,
+      "step": 341000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00039145932283750107,
+      "loss": 4.7653,
+      "step": 342000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00038084013217180266,
+      "loss": 4.7047,
+      "step": 343000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0003807297139737221,
+      "loss": 4.6748,
+      "step": 344000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0003806191054905468,
+      "loss": 4.6694,
+      "step": 345000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00038050808546821253,
+      "loss": 4.6753,
+      "step": 346000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0003803967646603707,
+      "loss": 4.6691,
+      "step": 347000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0003802851432525181,
+      "loss": 4.6667,
+      "step": 348000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0003801733335024691,
+      "loss": 4.6662,
+      "step": 349000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0003800611117532231,
+      "loss": 4.6614,
+      "step": 350000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000379948589963274,
+      "loss": 4.6566,
+      "step": 351000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00037983588129147694,
+      "loss": 4.6586,
+      "step": 352000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0003797227602826864,
+      "loss": 4.6598,
+      "step": 353000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00037960933979699685,
+      "loss": 4.6561,
+      "step": 354000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0003794956200234039,
+      "loss": 4.649,
+      "step": 355000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00037938171531961043,
+      "loss": 4.6508,
+      "step": 356000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0003792673978380055,
+      "loss": 4.6515,
+      "step": 357000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00037915301116867755,
+      "loss": 4.6502,
+      "step": 358000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00037903821210187236,
+      "loss": 4.6446,
+      "step": 359000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00037892299993410043,
+      "loss": 4.6457,
+      "step": 360000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0003788074896220918,
+      "loss": 4.6399,
+      "step": 361000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0003786916813583244,
+      "loss": 4.6416,
+      "step": 362000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0003785755753357728,
+      "loss": 4.6394,
+      "step": 363000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00037845917174790744,
+      "loss": 4.644,
+      "step": 364000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0003783425876381264,
+      "loss": 4.6455,
+      "step": 365000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0003782255897991082,
+      "loss": 4.6427,
+      "step": 366000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00037810841242106534,
+      "loss": 4.6383,
+      "step": 367000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003779908211099408,
+      "loss": 4.6398,
+      "step": 368000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003778729332078945,
+      "loss": 4.639,
+      "step": 369000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00037775474891136603,
+      "loss": 4.642,
+      "step": 370000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0003776363870456683,
+      "loss": 4.6378,
+      "step": 371000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00037751761084737167,
+      "loss": 4.6251,
+      "step": 372000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003773986580663642,
+      "loss": 4.6378,
+      "step": 373000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003772792907571875,
+      "loss": 4.638,
+      "step": 374000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003771598676628421,
+      "loss": 4.6329,
+      "step": 375000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0003770399103327158,
+      "loss": 4.6331,
+      "step": 376000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.000376919657996196,
+      "loss": 4.6307,
+      "step": 377000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0003767992315479937,
+      "loss": 4.6366,
+      "step": 378000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0003766783900948219,
+      "loss": 4.6312,
+      "step": 379000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0003765572542376675,
+      "loss": 4.6322,
+      "step": 380000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00037643582417838255,
+      "loss": 4.6272,
+      "step": 381000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0003763142219901536,
+      "loss": 4.6261,
+      "step": 382000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0003761922044278193,
+      "loss": 4.6332,
+      "step": 383000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0003760698932716468,
+      "loss": 4.6285,
+      "step": 384000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0003759472887254464,
+      "loss": 4.6315,
+      "step": 385000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00037582451403762754,
+      "loss": 4.6252,
+      "step": 386000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00037570132361763626,
+      "loss": 4.6238,
+      "step": 387000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00037557808768022013,
+      "loss": 4.6309,
+      "step": 388000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0003754543124991863,
+      "loss": 4.6227,
+      "step": 389000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0003753302449538835,
+      "loss": 4.6264,
+      "step": 390000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0003752058852510489,
+      "loss": 4.6297,
+      "step": 391000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00037508135839531953,
+      "loss": 4.6229,
+      "step": 392000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003749564152912182,
+      "loss": 4.6277,
+      "step": 393000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003748313060326983,
+      "loss": 4.6258,
+      "step": 394000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003747057803592816,
+      "loss": 4.6298,
+      "step": 395000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0003745799635688954,
+      "loss": 4.6275,
+      "step": 396000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0003744538558711915,
+      "loss": 4.6305,
+      "step": 397000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00037432758401983454,
+      "loss": 4.6254,
+      "step": 398000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0003742008954287709,
+      "loss": 4.6232,
+      "step": 399000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00037407404368583003,
+      "loss": 4.6243,
+      "step": 400000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0003739467750449806,
+      "loss": 4.6271,
+      "step": 401000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00037381921655191264,
+      "loss": 4.625,
+      "step": 402000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0003736914964119172,
+      "loss": 4.6207,
+      "step": 403000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0003735633591418774,
+      "loss": 4.6222,
+      "step": 404000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0003734349326585155,
+      "loss": 4.6274,
+      "step": 405000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00037330621717583185,
+      "loss": 4.6215,
+      "step": 406000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00037317734205675264,
+      "loss": 4.6239,
+      "step": 407000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00037304817894443345,
+      "loss": 4.6213,
+      "step": 408000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0003729185983290953,
+      "loss": 4.6217,
+      "step": 409000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00037278872957481737,
+      "loss": 4.6203,
+      "step": 410000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00037265870319842543,
+      "loss": 4.6233,
+      "step": 411000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0003725282591035563,
+      "loss": 4.6189,
+      "step": 412000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0003723976583952915,
+      "loss": 4.6208,
+      "step": 413000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0003722666398284116,
+      "loss": 4.6228,
+      "step": 414000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.000372135334208968,
+      "loss": 4.6152,
+      "step": 415000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00037200374175575874,
+      "loss": 4.6127,
+      "step": 416000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00037187199471021856,
+      "loss": 4.6182,
+      "step": 417000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0003717399618422258,
+      "loss": 4.6196,
+      "step": 418000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0003716075107774151,
+      "loss": 4.6225,
+      "step": 419000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00037147477375836516,
+      "loss": 4.6181,
+      "step": 420000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0003713418841716614,
+      "loss": 4.6207,
+      "step": 421000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00037120857619355976,
+      "loss": 4.6168,
+      "step": 422000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00037107511666167,
+      "loss": 4.6148,
+      "step": 423000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0003709412386121666,
+      "loss": 4.6241,
+      "step": 424000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00037080707571865136,
+      "loss": 4.6167,
+      "step": 425000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.000370672762794291,
+      "loss": 4.618,
+      "step": 426000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0003705380311681886,
+      "loss": 4.6185,
+      "step": 427000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00037040301536994983,
+      "loss": 4.6159,
+      "step": 428000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00037026771562455524,
+      "loss": 4.6172,
+      "step": 429000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0003701322678825694,
+      "loss": 4.6185,
+      "step": 430000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0003699964012030795,
+      "loss": 4.6142,
+      "step": 431000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0003698605238364365,
+      "loss": 4.619,
+      "step": 432000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0003697240914104684,
+      "loss": 4.6125,
+      "step": 433000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0003695873761686538,
+      "loss": 4.613,
+      "step": 434000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00036945037833880495,
+      "loss": 4.6193,
+      "step": 435000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00036931337299122744,
+      "loss": 4.6195,
+      "step": 436000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00036917581123466377,
+      "loss": 4.6155,
+      "step": 437000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0003690379675758677,
+      "loss": 4.6124,
+      "step": 438000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0003688998422445319,
+      "loss": 4.6118,
+      "step": 439000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0003687617125650919,
+      "loss": 4.6118,
+      "step": 440000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00036862302514182444,
+      "loss": 4.6115,
+      "step": 441000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0003684841958461244,
+      "loss": 4.6145,
+      "step": 442000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0003683449469728375,
+      "loss": 4.6107,
+      "step": 443000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00036820541758180987,
+      "loss": 4.6125,
+      "step": 444000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00036806574785514423,
+      "loss": 4.6076,
+      "step": 445000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0003679256584065426,
+      "loss": 4.6135,
+      "step": 446000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00036778528913887205,
+      "loss": 4.611,
+      "step": 447000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0003676447810744613,
+      "loss": 4.6169,
+      "step": 448000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00036750385315005585,
+      "loss": 4.6124,
+      "step": 449000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0003673626461094468,
+      "loss": 4.6091,
+      "step": 450000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00036722130181307566,
+      "loss": 4.6084,
+      "step": 451000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0003670795375249432,
+      "loss": 4.6093,
+      "step": 452000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0003669376370093399,
+      "loss": 4.6098,
+      "step": 453000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00036679531641764155,
+      "loss": 4.6088,
+      "step": 454000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00036665271789039375,
+      "loss": 4.6135,
+      "step": 455000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00036650984166521224,
+      "loss": 4.6111,
+      "step": 456000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00036636697456429214,
+      "loss": 4.6087,
+      "step": 457000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00036622354421214545,
+      "loss": 4.6105,
+      "step": 458000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0003660798368772088,
+      "loss": 4.6142,
+      "step": 459000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00036593599692117735,
+      "loss": 4.6036,
+      "step": 460000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00036579173661589563,
+      "loss": 4.6129,
+      "step": 461000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00036564720004735664,
+      "loss": 4.6066,
+      "step": 462000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00036550253240678936,
+      "loss": 4.6065,
+      "step": 463000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0003653574443103918,
+      "loss": 4.6065,
+      "step": 464000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0003652122261755973,
+      "loss": 4.6055,
+      "step": 465000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00036506658751743075,
+      "loss": 4.6077,
+      "step": 466000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0003649206738043425,
+      "loss": 4.6113,
+      "step": 467000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00036477463160518477,
+      "loss": 4.6056,
+      "step": 468000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00036462816878657725,
+      "loss": 4.608,
+      "step": 469000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.000364481431643597,
+      "loss": 4.6023,
+      "step": 470000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0003643345675688004,
+      "loss": 4.6027,
+      "step": 471000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00036418728278478005,
+      "loss": 4.6068,
+      "step": 472000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00036403972441104724,
+      "loss": 4.6072,
+      "step": 473000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0003638920406616534,
+      "loss": 4.6038,
+      "step": 474000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00036374393611956704,
+      "loss": 4.6014,
+      "step": 475000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00036359555872652883,
+      "loss": 4.6013,
+      "step": 476000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00036344705751586385,
+      "loss": 4.606,
+      "step": 477000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00024019467959966674,
+      "loss": 4.4837,
+      "step": 478000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00023963222037118084,
+      "loss": 4.4223,
+      "step": 479000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.00023906943468937218,
+      "loss": 4.39,
+      "step": 480000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00023850689045664867,
+      "loss": 4.3772,
+      "step": 481000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00023794346609281965,
+      "loss": 4.3726,
+      "step": 482000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00023737972918605284,
+      "loss": 4.3631,
+      "step": 483000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00023681624857694363,
+      "loss": 4.3497,
+      "step": 484000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00023625190081838816,
+      "loss": 4.3482,
+      "step": 485000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00023568781924440977,
+      "loss": 4.3443,
+      "step": 486000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0002351228792097228,
+      "loss": 4.3383,
+      "step": 487000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.000234558215237771,
+      "loss": 4.3393,
+      "step": 488000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0002339927015221048,
+      "loss": 4.3349,
+      "step": 489000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00023342690780622,
+      "loss": 4.3287,
+      "step": 490000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00023286083875059848,
+      "loss": 4.3263,
+      "step": 491000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002322950654913731,
+      "loss": 4.3202,
+      "step": 492000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002317284600104378,
+      "loss": 4.3198,
+      "step": 493000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.00023116216017581755,
+      "loss": 4.3163,
+      "step": 494000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00023059503691953928,
+      "loss": 4.3163,
+      "step": 495000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00023002766164983935,
+      "loss": 4.3123,
+      "step": 496000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00022946060678482666,
+      "loss": 4.3091,
+      "step": 497000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00022889274175117623,
+      "loss": 4.3067,
+      "step": 498000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00022832463872602635,
+      "loss": 4.3056,
+      "step": 499000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00022775687084019932,
+      "loss": 4.3042,
+      "step": 500000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00022718887477616112,
+      "loss": 4.306,
+      "step": 501000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00022662008630440305,
+      "loss": 4.3007,
+      "step": 502000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0002260510785611647,
+      "loss": 4.2996,
+      "step": 503000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00022548185623340192,
+      "loss": 4.2993,
+      "step": 504000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00022491299354534364,
+      "loss": 4.2962,
+      "step": 505000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0002243433563192932,
+      "loss": 4.2954,
+      "step": 506000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00022377408851168427,
+      "loss": 4.2964,
+      "step": 507000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00022320405513710757,
+      "loss": 4.2958,
+      "step": 508000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00022263440094754997,
+      "loss": 4.2934,
+      "step": 509000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0002220639901872479,
+      "loss": 4.2935,
+      "step": 510000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00022149396836606137,
+      "loss": 4.2906,
+      "step": 511000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00022092319899525643,
+      "loss": 4.2897,
+      "step": 512000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00022035282830486165,
+      "loss": 4.2916,
+      "step": 513000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00021978171911058022,
+      "loss": 4.2891,
+      "step": 514000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00021921101832488073,
+      "loss": 4.2867,
+      "step": 515000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00021863958810533452,
+      "loss": 4.288,
+      "step": 516000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00021806800435022003,
+      "loss": 4.2857,
+      "step": 517000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00021749684357306648,
+      "loss": 4.285,
+      "step": 518000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00021692496701433082,
+      "loss": 4.2832,
+      "step": 519000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00021635352312751783,
+      "loss": 4.2824,
+      "step": 520000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002157819448048862,
+      "loss": 4.2806,
+      "step": 521000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002152096644013863,
+      "loss": 4.2772,
+      "step": 522000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00021463725871483544,
+      "loss": 4.2798,
+      "step": 523000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0002140647324601787,
+      "loss": 4.2798,
+      "step": 524000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00021349266305175916,
+      "loss": 4.2802,
+      "step": 525000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00021291990991843793,
+      "loss": 4.2786,
+      "step": 526000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002123476232740738,
+      "loss": 4.2791,
+      "step": 527000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00021177466211441055,
+      "loss": 4.274,
+      "step": 528000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002112021770710695,
+      "loss": 4.2765,
+      "step": 529000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00021062959993907988,
+      "loss": 4.2751,
+      "step": 530000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00021005636214541413,
+      "loss": 4.2751,
+      "step": 531000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00020948304151680226,
+      "loss": 4.2744,
+      "step": 532000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00020891021621191204,
+      "loss": 4.273,
+      "step": 533000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00020833674415252564,
+      "loss": 4.2769,
+      "step": 534000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00020776320342280467,
+      "loss": 4.2695,
+      "step": 535000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00020718959874704363,
+      "loss": 4.2689,
+      "step": 536000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00020661650854196894,
+      "loss": 4.2671,
+      "step": 537000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002060427902012143,
+      "loss": 4.2677,
+      "step": 538000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002054701696683469,
+      "loss": 4.2713,
+      "step": 539000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00020489635658938387,
+      "loss": 4.2659,
+      "step": 540000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0002043225031787951,
+      "loss": 4.2666,
+      "step": 541000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00020374861416345058,
+      "loss": 4.2609,
+      "step": 542000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0002031746942705136,
+      "loss": 4.2631,
+      "step": 543000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0002026013221849334,
+      "loss": 4.2656,
+      "step": 544000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0002020273547383406,
+      "loss": 4.2688,
+      "step": 545000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00020145394458320146,
+      "loss": 4.2592,
+      "step": 546000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00020087994847524482,
+      "loss": 4.2617,
+      "step": 547000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00020030651912449513,
+      "loss": 4.2613,
+      "step": 548000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00019973251324840986,
+      "loss": 4.2557,
+      "step": 549000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0001991590835766299,
+      "loss": 4.2604,
+      "step": 550000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00019858508682597277,
+      "loss": 4.2614,
+      "step": 551000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00019801167570775345,
+      "loss": 4.2587,
+      "step": 552000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0001974377069757808,
+      "loss": 4.2567,
+      "step": 553000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.000196863759349592,
+      "loss": 4.2542,
+      "step": 554000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0001962904114641484,
+      "loss": 4.2528,
+      "step": 555000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00019571652019933017,
+      "loss": 4.2529,
+      "step": 556000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00019514323805461362,
+      "loss": 4.2504,
+      "step": 557000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00019456999583540802,
+      "loss": 4.2557,
+      "step": 558000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00019399622450669583,
+      "loss": 4.2527,
+      "step": 559000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00019342250263149486,
+      "loss": 4.2493,
+      "step": 560000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0001928488349355918,
+      "loss": 4.2533,
+      "step": 561000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00019227579972212256,
+      "loss": 4.251,
+      "step": 562000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00019170225449436132,
+      "loss": 4.2442,
+      "step": 563000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00019112935105686604,
+      "loss": 4.2508,
+      "step": 564000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00019055652054145262,
+      "loss": 4.2482,
+      "step": 565000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00018998319437138936,
+      "loss": 4.2453,
+      "step": 566000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0001894099507104425,
+      "loss": 4.2467,
+      "step": 567000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00018883679428045936,
+      "loss": 4.2429,
+      "step": 568000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00018826430281954561,
+      "loss": 4.2436,
+      "step": 569000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00018769190783313742,
+      "loss": 4.2462,
+      "step": 570000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00018711904121225677,
+      "loss": 4.2429,
+      "step": 571000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.000186546280692719,
+      "loss": 4.2415,
+      "step": 572000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0001859736309923917,
+      "loss": 4.242,
+      "step": 573000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00018540166930311399,
+      "loss": 4.2415,
+      "step": 574000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00018482925526851332,
+      "loss": 4.2394,
+      "step": 575000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00018425696619637965,
+      "loss": 4.2393,
+      "step": 576000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00018368537889375085,
+      "loss": 4.2374,
+      "step": 577000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00018311335375069304,
+      "loss": 4.2376,
+      "step": 578000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00018254203951910075,
+      "loss": 4.2361,
+      "step": 579000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00018197029713347917,
+      "loss": 4.2363,
+      "step": 580000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.000181399274777884,
+      "loss": 4.2322,
+      "step": 581000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00018082783396875207,
+      "loss": 4.235,
+      "step": 582000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00018025655108206925,
+      "loss": 4.2327,
+      "step": 583000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0001796854308235321,
+      "loss": 4.2323,
+      "step": 584000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0001791150487652753,
+      "loss": 4.2297,
+      "step": 585000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00017854426770033718,
+      "loss": 4.2339,
+      "step": 586000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00017797423388223084,
+      "loss": 4.2315,
+      "step": 587000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00017740381079830306,
+      "loss": 4.2289,
+      "step": 588000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0001768341439831626,
+      "loss": 4.2285,
+      "step": 589000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00017626409765587338,
+      "loss": 4.2273,
+      "step": 590000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.000175694246842843,
+      "loss": 4.2272,
+      "step": 591000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00017512459623797167,
+      "loss": 4.2267,
+      "step": 592000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00017455571987530613,
+      "loss": 4.2242,
+      "step": 593000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00017398648354988546,
+      "loss": 4.2238,
+      "step": 594000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00017341803041304732,
+      "loss": 4.2245,
+      "step": 595000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00017284922710364303,
+      "loss": 4.2219,
+      "step": 596000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00017228121590341918,
+      "loss": 4.2215,
+      "step": 597000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0001717128643323442,
+      "loss": 4.2196,
+      "step": 598000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00017114474576434977,
+      "loss": 4.2186,
+      "step": 599000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00017057686487906743,
+      "loss": 4.2218,
+      "step": 600000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00017000922635417116,
+      "loss": 4.2175,
+      "step": 601000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00016944183486533842,
+      "loss": 4.2174,
+      "step": 602000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00016887582911145858,
+      "loss": 4.2208,
+      "step": 603000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00016830894519618436,
+      "loss": 4.2176,
+      "step": 604000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00016774232232230643,
+      "loss": 4.2131,
+      "step": 605000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00016717596515713635,
+      "loss": 4.2148,
+      "step": 606000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00016661044431598456,
+      "loss": 4.2163,
+      "step": 607000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0001660446322840068,
+      "loss": 4.2121,
+      "step": 608000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0001654796653358085,
+      "loss": 4.2126,
+      "step": 609000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0001649144170608772,
+      "loss": 4.2111,
+      "step": 610000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00016435002260167044,
+      "loss": 4.2093,
+      "step": 611000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.0001637853566890836,
+      "loss": 4.2104,
+      "step": 612000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00016322155329606282,
+      "loss": 4.2104,
+      "step": 613000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00016265748833194975,
+      "loss": 4.2095,
+      "step": 614000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00016209373096067142,
+      "loss": 4.2061,
+      "step": 615000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.00016153141240150847,
+      "loss": 4.2059,
+      "step": 616000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.0001609682835060673,
+      "loss": 4.2093,
+      "step": 617000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00016040547611755718,
+      "loss": 4.2025,
+      "step": 618000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00015984299487186134,
+      "loss": 4.2069,
+      "step": 619000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00015928140638588216,
+      "loss": 4.2031,
+      "step": 620000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00015872015263128903,
+      "loss": 4.2021,
+      "step": 621000
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00015815867691759442,
+      "loss": 4.2014,
+      "step": 622000
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00015759754585375357,
+      "loss": 4.2014,
+      "step": 623000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00015703676406184148,
+      "loss": 4.2015,
+      "step": 624000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0001564768964106519,
+      "loss": 4.2018,
+      "step": 625000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.000155917386545611,
+      "loss": 4.2003,
+      "step": 626000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00015535767954213264,
+      "loss": 4.1976,
+      "step": 627000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00015479834026051583,
+      "loss": 4.1972,
+      "step": 628000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00015423937330807675,
+      "loss": 4.1957,
+      "step": 629000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00015368134168927352,
+      "loss": 4.1951,
+      "step": 630000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00015312313282100077,
+      "loss": 4.1952,
+      "step": 631000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00015256586770904422,
+      "loss": 4.1928,
+      "step": 632000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00015200843529853173,
+      "loss": 4.1941,
+      "step": 633000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00015145195503595184,
+      "loss": 4.1938,
+      "step": 634000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00015089531743123636,
+      "loss": 4.1933,
+      "step": 635000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00015033964033472967,
+      "loss": 4.1919,
+      "step": 636000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00014978381585768676,
+      "loss": 4.191,
+      "step": 637000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.0001492289602175133,
+      "loss": 4.1907,
+      "step": 638000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00014867396716325404,
+      "loss": 4.1906,
+      "step": 639000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00014811995124263547,
+      "loss": 4.19,
+      "step": 640000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00014756580787890456,
+      "loss": 4.1854,
+      "step": 641000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0001470126499134229,
+      "loss": 4.1862,
+      "step": 642000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0001464593744799972,
+      "loss": 4.1804,
+      "step": 643000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00014590709267699477,
+      "loss": 4.1812,
+      "step": 644000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00014535470338508303,
+      "loss": 4.1811,
+      "step": 645000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.0001448033159230627,
+      "loss": 4.1812,
+      "step": 646000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00014425238221106002,
+      "loss": 4.1827,
+      "step": 647000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00014370135598273356,
+      "loss": 4.1792,
+      "step": 648000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00014315079349020695,
+      "loss": 4.1829,
+      "step": 649000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00014260069926850117,
+      "loss": 4.1798,
+      "step": 650000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00014205162723252818,
+      "loss": 4.1813,
+      "step": 651000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00014150248266247203,
+      "loss": 4.1771,
+      "step": 652000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.0001409543683610207,
+      "loss": 4.1744,
+      "step": 653000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00014040673945025616,
+      "loss": 4.1791,
+      "step": 654000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00013985905299225343,
+      "loss": 4.1795,
+      "step": 655000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00013931186191936434,
+      "loss": 4.1764,
+      "step": 656000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0001387651707388392,
+      "loss": 4.1717,
+      "step": 657000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0001382195298871527,
+      "loss": 4.1585,
+      "step": 658000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00013767385148545907,
+      "loss": 4.1503,
+      "step": 659000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.0001371292313756203,
+      "loss": 4.1517,
+      "step": 660000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00013658458371390849,
+      "loss": 4.1508,
+      "step": 661000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00013604100227223385,
+      "loss": 4.1545,
+      "step": 662000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00013549740327772723,
+      "loss": 4.1505,
+      "step": 663000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00013495542119768334,
+      "loss": 4.1496,
+      "step": 664000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00013441288822507396,
+      "loss": 4.1504,
+      "step": 665000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.0001338708954980116,
+      "loss": 4.1514,
+      "step": 666000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.0001333299886553773,
+      "loss": 4.1504,
+      "step": 667000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00013278908925682,
+      "loss": 4.1553,
+      "step": 668000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0001322487434791535,
+      "loss": 4.154,
+      "step": 669000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00013170895577324293,
+      "loss": 4.1501,
+      "step": 670000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00013117026952808839,
+      "loss": 4.1481,
+      "step": 671000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00013063161073068494,
+      "loss": 4.153,
+      "step": 672000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00013009406112599048,
+      "loss": 4.148,
+      "step": 673000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00012955708619025508,
+      "loss": 4.1457,
+      "step": 674000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00012902015369654687,
+      "loss": 4.1496,
+      "step": 675000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00012848434192302686,
+      "loss": 4.1481,
+      "step": 676000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00012794858258770753,
+      "loss": 4.148,
+      "step": 677000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00012741341674486485,
+      "loss": 4.1484,
+      "step": 678000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00012687884880269694,
+      "loss": 4.1446,
+      "step": 679000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00012634541682779958,
+      "loss": 4.1428,
+      "step": 680000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00012581205728294073,
+      "loss": 4.1455,
+      "step": 681000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00012527984127101713,
+      "loss": 4.1422,
+      "step": 682000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.0001247477076791393,
+      "loss": 4.1427,
+      "step": 683000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00012421672514822168,
+      "loss": 4.1434,
+      "step": 684000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00012368583502464424,
+      "loss": 4.1414,
+      "step": 685000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00012315610345216445,
+      "loss": 4.1437,
+      "step": 686000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00012262647427127763,
+      "loss": 4.1419,
+      "step": 687000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0001220980110934919,
+      "loss": 4.1379,
+      "step": 688000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00012156966028818173,
+      "loss": 4.1382,
+      "step": 689000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00012104248289959676,
+      "loss": 4.1365,
+      "step": 690000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00012051542786067112,
+      "loss": 4.1394,
+      "step": 691000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00011998955361347148,
+      "loss": 4.1366,
+      "step": 692000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00011946381168908787,
+      "loss": 4.1347,
+      "step": 693000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00011893873314682198,
+      "loss": 4.1357,
+      "step": 694000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00011841432231178195,
+      "loss": 4.1337,
+      "step": 695000
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.0001178911069052703,
+      "loss": 4.1347,
+      "step": 696000
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00011736804375947676,
+      "loss": 4.1351,
+      "step": 697000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00011684618329987129,
+      "loss": 4.1297,
+      "step": 698000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00011632448506008744,
+      "loss": 4.1351,
+      "step": 699000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00011580399672456457,
+      "loss": 4.1329,
+      "step": 700000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00011528368056262728,
+      "loss": 4.1313,
+      "step": 701000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00011476458148319966,
+      "loss": 4.1265,
+      "step": 702000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00011424566452545455,
+      "loss": 4.129,
+      "step": 703000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00011372797178840713,
+      "loss": 4.1299,
+      "step": 704000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00011321047111514422,
+      "loss": 4.1257,
+      "step": 705000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.0001126942017604717,
+      "loss": 4.1281,
+      "step": 706000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00011217813440536418,
+      "loss": 4.1266,
+      "step": 707000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00011166279044499894,
+      "loss": 4.1249,
+      "step": 708000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00011114817412429949,
+      "loss": 4.1247,
+      "step": 709000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00011063428968219605,
+      "loss": 4.1229,
+      "step": 710000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00011012114135158998,
+      "loss": 4.1245,
+      "step": 711000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010960924539610728,
+      "loss": 4.1261,
+      "step": 712000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010909758121624652,
+      "loss": 4.1228,
+      "step": 713000
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010858717634585534,
+      "loss": 4.1197,
+      "step": 714000
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010807701315830314,
+      "loss": 4.1174,
+      "step": 715000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00010756862520028245,
+      "loss": 4.1188,
+      "step": 716000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00010705997903373485,
+      "loss": 4.1191,
+      "step": 717000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00010655209842052723,
+      "loss": 4.114,
+      "step": 718000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00010604549426910888,
+      "loss": 4.1175,
+      "step": 719000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010553915653058473,
+      "loss": 4.1166,
+      "step": 720000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010503359687251983,
+      "loss": 4.1143,
+      "step": 721000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010452881945924391,
+      "loss": 4.1152,
+      "step": 722000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00010402533204546334,
+      "loss": 4.1116,
+      "step": 723000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00010352213079632074,
+      "loss": 4.1111,
+      "step": 724000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00010301972424201705,
+      "loss": 4.1103,
+      "step": 725000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00010251861772823774,
+      "loss": 4.1111,
+      "step": 726000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00010201781216707713,
+      "loss": 4.1106,
+      "step": 727000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00010151831328589558,
+      "loss": 4.1069,
+      "step": 728000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010101962398354699,
+      "loss": 4.1067,
+      "step": 729000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010052174835955799,
+      "loss": 4.1035,
+      "step": 730000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00010002419336242872,
+      "loss": 4.1095,
+      "step": 731000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 9.952746187288931e-05,
+      "loss": 4.1049,
+      "step": 732000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.903155798255135e-05,
+      "loss": 4.0988,
+      "step": 733000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.853648577620898e-05,
+      "loss": 4.1043,
+      "step": 734000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.804274314943199e-05,
+      "loss": 4.1043,
+      "step": 735000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 9.754934569616405e-05,
+      "loss": 4.1018,
+      "step": 736000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 9.705777639819362e-05,
+      "loss": 4.099,
+      "step": 737000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.656606908833878e-05,
+      "loss": 4.1011,
+      "step": 738000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.6075213770881e-05,
+      "loss": 4.1025,
+      "step": 739000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.558570405937759e-05,
+      "loss": 4.1005,
+      "step": 740000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.509656398720454e-05,
+      "loss": 4.0979,
+      "step": 741000
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.46082880118432e-05,
+      "loss": 4.0973,
+      "step": 742000
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.412088015525628e-05,
+      "loss": 4.0912,
+      "step": 743000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.363434443225589e-05,
+      "loss": 4.0913,
+      "step": 744000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.31486848504702e-05,
+      "loss": 4.0922,
+      "step": 745000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.266390541031052e-05,
+      "loss": 4.095,
+      "step": 746000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.218049355729118e-05,
+      "loss": 4.0916,
+      "step": 747000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.169748548247643e-05,
+      "loss": 4.0896,
+      "step": 748000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.121585117197211e-05,
+      "loss": 4.0896,
+      "step": 749000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.073463036084202e-05,
+      "loss": 4.0849,
+      "step": 750000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 9.025430957607068e-05,
+      "loss": 4.0861,
+      "step": 751000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 8.977489277409341e-05,
+      "loss": 4.0873,
+      "step": 752000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.929686195794506e-05,
+      "loss": 4.083,
+      "step": 753000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 8.882021833036489e-05,
+      "loss": 4.0873,
+      "step": 754000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.834353438745977e-05,
+      "loss": 4.083,
+      "step": 755000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.786824547005008e-05,
+      "loss": 4.0832,
+      "step": 756000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 8.739340396441291e-05,
+      "loss": 4.0828,
+      "step": 757000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.691949000704588e-05,
+      "loss": 4.087,
+      "step": 758000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 8.644650750161096e-05,
+      "loss": 4.0797,
+      "step": 759000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.597446034409749e-05,
+      "loss": 4.0808,
+      "step": 760000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 8.55033524227903e-05,
+      "loss": 4.0762,
+      "step": 761000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 8.503365731066581e-05,
+      "loss": 4.0769,
+      "step": 762000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 8.456443854672643e-05,
+      "loss": 4.0744,
+      "step": 763000
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.409617063343962e-05,
+      "loss": 4.0785,
+      "step": 764000
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 8.362885742796067e-05,
+      "loss": 4.074,
+      "step": 765000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.316296865415034e-05,
+      "loss": 4.0752,
+      "step": 766000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.269757543994949e-05,
+      "loss": 4.0721,
+      "step": 767000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 8.223314845388103e-05,
+      "loss": 4.0745,
+      "step": 768000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.176969152146221e-05,
+      "loss": 4.0721,
+      "step": 769000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 8.130767045556329e-05,
+      "loss": 4.0721,
+      "step": 770000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.084616409542043e-05,
+      "loss": 4.0681,
+      "step": 771000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 8.038609924698259e-05,
+      "loss": 4.0728,
+      "step": 772000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.992701769691633e-05,
+      "loss": 4.0687,
+      "step": 773000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.946846516190165e-05,
+      "loss": 4.0675,
+      "step": 774000
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 7.90109054534227e-05,
+      "loss": 4.0665,
+      "step": 775000
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 7.855434234043022e-05,
+      "loss": 4.0655,
+      "step": 776000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.80992346454953e-05,
+      "loss": 4.0671,
+      "step": 777000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.764512904833741e-05,
+      "loss": 4.0638,
+      "step": 778000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 7.71915762338268e-05,
+      "loss": 4.0613,
+      "step": 779000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 7.673948703544935e-05,
+      "loss": 4.0572,
+      "step": 780000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 7.628796009423646e-05,
+      "loss": 4.0639,
+      "step": 781000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 7.583745217759814e-05,
+      "loss": 4.0613,
+      "step": 782000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 7.538841596949084e-05,
+      "loss": 4.0572,
+      "step": 783000
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 7.493995619788687e-05,
+      "loss": 4.0566,
+      "step": 784000
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 7.44925265544582e-05,
+      "loss": 4.0562,
+      "step": 785000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 7.404613072471351e-05,
+      "loss": 4.056,
+      "step": 786000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 7.360077238564593e-05,
+      "loss": 4.0507,
+      "step": 787000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.315645520570287e-05,
+      "loss": 4.0505,
+      "step": 788000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.271362559401307e-05,
+      "loss": 4.0534,
+      "step": 789000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 7.22714006530347e-05,
+      "loss": 4.0539,
+      "step": 790000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 7.183066846739989e-05,
+      "loss": 4.0536,
+      "step": 791000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 7.139055032133843e-05,
+      "loss": 4.0522,
+      "step": 792000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 7.095236859806331e-05,
+      "loss": 4.05,
+      "step": 793000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 7.051437066874354e-05,
+      "loss": 4.0474,
+      "step": 794000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 7.007743932145127e-05,
+      "loss": 4.0424,
+      "step": 795000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 6.964157815522e-05,
+      "loss": 4.0445,
+      "step": 796000
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 6.920679076026799e-05,
+      "loss": 4.0437,
+      "step": 797000
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 6.877308071796904e-05,
+      "loss": 4.0428,
+      "step": 798000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 6.834088368883074e-05,
+      "loss": 4.0413,
+      "step": 799000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 6.790933797416663e-05,
+      "loss": 4.0423,
+      "step": 800000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 6.747888029936322e-05,
+      "loss": 4.0412,
+      "step": 801000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 6.704994302979443e-05,
+      "loss": 4.0394,
+      "step": 802000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 6.66216709659637e-05,
+      "loss": 4.0399,
+      "step": 803000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 6.619449754859523e-05,
+      "loss": 4.0385,
+      "step": 804000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 6.576885181589794e-05,
+      "loss": 4.0375,
+      "step": 805000
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 6.534388513092143e-05,
+      "loss": 4.0376,
+      "step": 806000
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 6.492002761761704e-05,
+      "loss": 4.0367,
+      "step": 807000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 6.44972827673282e-05,
+      "loss": 4.0361,
+      "step": 808000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 6.40760751322673e-05,
+      "loss": 4.0294,
+      "step": 809000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 6.365556492400127e-05,
+      "loss": 4.0324,
+      "step": 810000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 6.323659661921848e-05,
+      "loss": 4.0314,
+      "step": 811000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 6.281833489418096e-05,
+      "loss": 4.0306,
+      "step": 812000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 6.240161971012996e-05,
+      "loss": 4.0301,
+      "step": 813000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 6.198562023551751e-05,
+      "loss": 4.0285,
+      "step": 814000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 6.157158618328416e-05,
+      "loss": 4.0266,
+      "step": 815000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 6.115786150987899e-05,
+      "loss": 4.0292,
+      "step": 816000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 6.0745280488710155e-05,
+      "loss": 4.0234,
+      "step": 817000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 6.0333846518236035e-05,
+      "loss": 4.0236,
+      "step": 818000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.9923562987466307e-05,
+      "loss": 4.0237,
+      "step": 819000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 5.951484182819116e-05,
+      "loss": 4.021,
+      "step": 820000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.910727554160531e-05,
+      "loss": 4.0171,
+      "step": 821000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 5.8700461244659956e-05,
+      "loss": 4.0197,
+      "step": 822000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.829481084172575e-05,
+      "loss": 4.0153,
+      "step": 823000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.789032767417306e-05,
+      "loss": 4.0209,
+      "step": 824000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 5.748701507375753e-05,
+      "loss": 4.0169,
+      "step": 825000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.708487636259276e-05,
+      "loss": 4.0142,
+      "step": 826000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 5.6684315225520025e-05,
+      "loss": 4.0168,
+      "step": 827000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5.628453303834178e-05,
+      "loss": 4.0179,
+      "step": 828000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 5.588633265133554e-05,
+      "loss": 4.0114,
+      "step": 829000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5489316965551574e-05,
+      "loss": 4.0113,
+      "step": 830000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 5.5093093617013605e-05,
+      "loss": 4.0101,
+      "step": 831000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5.469806387662206e-05,
+      "loss": 4.012,
+      "step": 832000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 5.4304230998263825e-05,
+      "loss": 4.011,
+      "step": 833000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.391199025820963e-05,
+      "loss": 4.0085,
+      "step": 834000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 5.352055962116598e-05,
+      "loss": 4.0103,
+      "step": 835000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.313033554533935e-05,
+      "loss": 4.007,
+      "step": 836000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.2741321245032015e-05,
+      "loss": 4.0057,
+      "step": 837000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 5.235429431454388e-05,
+      "loss": 4.0023,
+      "step": 838000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.196770673276694e-05,
+      "loss": 4.0024,
+      "step": 839000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 5.158233850316285e-05,
+      "loss": 4.0052,
+      "step": 840000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.119895986925622e-05,
+      "loss": 3.9984,
+      "step": 841000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 5.0816037402308914e-05,
+      "loss": 4.0024,
+      "step": 842000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5.0434343773913936e-05,
+      "loss": 3.9997,
+      "step": 843000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 5.005388212810789e-05,
+      "loss": 4.0001,
+      "step": 844000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.967465559877949e-05,
+      "loss": 3.9953,
+      "step": 845000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.929666730964366e-05,
+      "loss": 3.9974,
+      "step": 846000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.8920296500061624e-05,
+      "loss": 3.9978,
+      "step": 847000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.854479277562882e-05,
+      "loss": 3.9965,
+      "step": 848000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.8170910230147306e-05,
+      "loss": 3.9938,
+      "step": 849000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.77979034302229e-05,
+      "loss": 3.9955,
+      "step": 850000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.7426521455285876e-05,
+      "loss": 3.9919,
+      "step": 851000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.705602385748844e-05,
+      "loss": 3.9902,
+      "step": 852000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.668678606973318e-05,
+      "loss": 3.9888,
+      "step": 853000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.631881113345728e-05,
+      "loss": 3.9864,
+      "step": 854000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.5952834232442806e-05,
+      "loss": 3.9886,
+      "step": 855000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.5587391540988944e-05,
+      "loss": 3.9877,
+      "step": 856000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.5223220756802585e-05,
+      "loss": 3.9858,
+      "step": 857000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.4860324879583624e-05,
+      "loss": 3.9823,
+      "step": 858000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.4498706898530285e-05,
+      "loss": 3.9823,
+      "step": 859000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.413836979231471e-05,
+      "loss": 3.9826,
+      "step": 860000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.3779674940056856e-05,
+      "loss": 3.9845,
+      "step": 861000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.342190718903205e-05,
+      "loss": 3.9797,
+      "step": 862000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.3066140849412765e-05,
+      "loss": 3.9826,
+      "step": 863000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.271095293545859e-05,
+      "loss": 3.9786,
+      "step": 864000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.235706062219449e-05,
+      "loss": 3.9789,
+      "step": 865000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.200481876887719e-05,
+      "loss": 3.9804,
+      "step": 866000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.165352508853595e-05,
+      "loss": 3.9759,
+      "step": 867000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.13038850558964e-05,
+      "loss": 3.9766,
+      "step": 868000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.095520157140329e-05,
+      "loss": 3.9754,
+      "step": 869000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.0608174867936735e-05,
+      "loss": 3.9726,
+      "step": 870000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.026211305630183e-05,
+      "loss": 3.9728,
+      "step": 871000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.9917367016619276e-05,
+      "loss": 3.9719,
+      "step": 872000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.9573939588586015e-05,
+      "loss": 3.9689,
+      "step": 873000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.923183360103733e-05,
+      "loss": 3.9669,
+      "step": 874000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 3.88910518719237e-05,
+      "loss": 3.9705,
+      "step": 875000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.8551935999150546e-05,
+      "loss": 3.969,
+      "step": 876000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.821414732678987e-05,
+      "loss": 3.9472,
+      "step": 877000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.7877352503423325e-05,
+      "loss": 3.9284,
+      "step": 878000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.7541893095445734e-05,
+      "loss": 3.9269,
+      "step": 879000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.720810531795154e-05,
+      "loss": 3.9298,
+      "step": 880000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.687532367703408e-05,
+      "loss": 3.9306,
+      "step": 881000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.65438857052858e-05,
+      "loss": 3.9292,
+      "step": 882000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.6213794132784204e-05,
+      "loss": 3.9297,
+      "step": 883000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.588537974618371e-05,
+      "loss": 3.9289,
+      "step": 884000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.555798776484851e-05,
+      "loss": 3.9268,
+      "step": 885000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.5232601024993396e-05,
+      "loss": 3.933,
+      "step": 886000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.490791805247826e-05,
+      "loss": 3.9279,
+      "step": 887000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.458459495478781e-05,
+      "loss": 3.9265,
+      "step": 888000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.4262634395156536e-05,
+      "loss": 3.9279,
+      "step": 889000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.394235893817297e-05,
+      "loss": 3.9273,
+      "step": 890000
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.3623130030302484e-05,
+      "loss": 3.924,
+      "step": 891000
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.330527158014394e-05,
+      "loss": 3.9239,
+      "step": 892000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.298910200457324e-05,
+      "loss": 3.9256,
+      "step": 893000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.267399093621268e-05,
+      "loss": 3.9225,
+      "step": 894000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.2360258143687926e-05,
+      "loss": 3.9213,
+      "step": 895000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.204821787257311e-05,
+      "loss": 3.923,
+      "step": 896000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.173724798834707e-05,
+      "loss": 3.9165,
+      "step": 897000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.142797298671269e-05,
+      "loss": 3.9243,
+      "step": 898000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.111977624644229e-05,
+      "loss": 3.9195,
+      "step": 899000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.081297058418091e-05,
+      "loss": 3.9208,
+      "step": 900000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.0507558527107828e-05,
+      "loss": 3.92,
+      "step": 901000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.0203845908662563e-05,
+      "loss": 3.9204,
+      "step": 902000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 2.9901227197694415e-05,
+      "loss": 3.9196,
+      "step": 903000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9600310118919393e-05,
+      "loss": 3.9197,
+      "step": 904000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9300494714831896e-05,
+      "loss": 3.9205,
+      "step": 905000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 2.9002383080493055e-05,
+      "loss": 3.9151,
+      "step": 906000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.8705380851790375e-05,
+      "loss": 3.9172,
+      "step": 907000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 2.8409789588637402e-05,
+      "loss": 3.9158,
+      "step": 908000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.8115611725839808e-05,
+      "loss": 3.9135,
+      "step": 909000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 2.7823141740592663e-05,
+      "loss": 3.9149,
+      "step": 910000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.7531796516897657e-05,
+      "loss": 3.9118,
+      "step": 911000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.7242161139836732e-05,
+      "loss": 3.9082,
+      "step": 912000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.6953658144950188e-05,
+      "loss": 3.9131,
+      "step": 913000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 2.666686690950142e-05,
+      "loss": 3.9105,
+      "step": 914000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.6381215639576494e-05,
+      "loss": 3.9091,
+      "step": 915000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.6097277985549907e-05,
+      "loss": 3.9072,
+      "step": 916000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 2.581476991673275e-05,
+      "loss": 3.9109,
+      "step": 917000
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.553341311615387e-05,
+      "loss": 3.9079,
+      "step": 918000
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 2.5253493408841024e-05,
+      "loss": 3.9061,
+      "step": 919000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.4975290861076127e-05,
+      "loss": 3.9058,
+      "step": 920000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 2.469825080275776e-05,
+      "loss": 3.9049,
+      "step": 921000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.4422929591059718e-05,
+      "loss": 3.9037,
+      "step": 922000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 2.4148778300583463e-05,
+      "loss": 3.9052,
+      "step": 923000
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.3876075508705364e-05,
+      "loss": 3.906,
+      "step": 924000
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 2.3605364516460604e-05,
+      "loss": 3.9016,
+      "step": 925000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.3335562540463497e-05,
+      "loss": 3.902,
+      "step": 926000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 2.3067215761578686e-05,
+      "loss": 3.901,
+      "step": 927000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.2800326390197003e-05,
+      "loss": 3.9034,
+      "step": 928000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.2535161324668153e-05,
+      "loss": 3.8986,
+      "step": 929000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 2.227119188854776e-05,
+      "loss": 3.9001,
+      "step": 930000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.200868641683378e-05,
+      "loss": 3.896,
+      "step": 931000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 2.1747647071801923e-05,
+      "loss": 3.8955,
+      "step": 932000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.1488593679023983e-05,
+      "loss": 3.896,
+      "step": 933000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.1230490082903298e-05,
+      "loss": 3.8937,
+      "step": 934000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.0973859023521336e-05,
+      "loss": 3.893,
+      "step": 935000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 2.0718957033886022e-05,
+      "loss": 3.8921,
+      "step": 936000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.0465275899699664e-05,
+      "loss": 3.8936,
+      "step": 937000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 2.021307360537388e-05,
+      "loss": 3.8894,
+      "step": 938000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.9962352228316283e-05,
+      "loss": 3.8913,
+      "step": 939000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.9713362330696583e-05,
+      "loss": 3.8938,
+      "step": 940000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.946560748553077e-05,
+      "loss": 3.8904,
+      "step": 941000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.921958523886409e-05,
+      "loss": 3.8881,
+      "step": 942000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.8974805080506908e-05,
+      "loss": 3.8859,
+      "step": 943000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.8732001127734854e-05,
+      "loss": 3.8898,
+      "step": 944000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.8490202219074714e-05,
+      "loss": 3.888,
+      "step": 945000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.8250137975426186e-05,
+      "loss": 3.8873,
+      "step": 946000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.8011329765448747e-05,
+      "loss": 3.8839,
+      "step": 947000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.7774020608654827e-05,
+      "loss": 3.8851,
+      "step": 948000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.753821245977625e-05,
+      "loss": 3.8815,
+      "step": 949000
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.730414081501248e-05,
+      "loss": 3.8797,
+      "step": 950000
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.7071338990848274e-05,
+      "loss": 3.8825,
+      "step": 951000
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.6840274504384723e-05,
+      "loss": 3.8792,
+      "step": 952000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.6610715701279632e-05,
+      "loss": 3.8822,
+      "step": 953000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.6382436945055167e-05,
+      "loss": 3.8788,
+      "step": 954000
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.615567065931629e-05,
+      "loss": 3.8825,
+      "step": 955000
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.5930643206869322e-05,
+      "loss": 3.8789,
+      "step": 956000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.5706905936180028e-05,
+      "loss": 3.8776,
+      "step": 957000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.5484908160738844e-05,
+      "loss": 3.88,
+      "step": 958000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.5264207269471153e-05,
+      "loss": 3.8763,
+      "step": 959000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.5045028059623756e-05,
+      "loss": 3.8744,
+      "step": 960000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.4827372336590928e-05,
+      "loss": 3.8755,
+      "step": 961000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.4611457261190308e-05,
+      "loss": 3.8775,
+      "step": 962000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.4397066191369536e-05,
+      "loss": 3.8748,
+      "step": 963000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.4183988576170026e-05,
+      "loss": 3.8712,
+      "step": 964000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.3972441540226522e-05,
+      "loss": 3.8697,
+      "step": 965000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.3762426826066322e-05,
+      "loss": 3.8722,
+      "step": 966000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.3554153877426224e-05,
+      "loss": 3.8705,
+      "step": 967000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.3347207447291144e-05,
+      "loss": 3.8681,
+      "step": 968000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.314179848903565e-05,
+      "loss": 3.8657,
+      "step": 969000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.2938131795049502e-05,
+      "loss": 3.8674,
+      "step": 970000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.2735801302100369e-05,
+      "loss": 3.8668,
+      "step": 971000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.2535213334175821e-05,
+      "loss": 3.8706,
+      "step": 972000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.2335967966295303e-05,
+      "loss": 3.8628,
+      "step": 973000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.2138268399943431e-05,
+      "loss": 3.8666,
+      "step": 974000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.1942116263585212e-05,
+      "loss": 3.8637,
+      "step": 975000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.1747707001746943e-05,
+      "loss": 3.8626,
+      "step": 976000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.1554653008327055e-05,
+      "loss": 3.8644,
+      "step": 977000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.1363532705509805e-05,
+      "loss": 3.8628,
+      "step": 978000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.1173581654855314e-05,
+      "loss": 3.864,
+      "step": 979000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.0985185980385471e-05,
+      "loss": 3.8634,
+      "step": 980000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.079834723392832e-05,
+      "loss": 3.8611,
+      "step": 981000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.0613066954487539e-05,
+      "loss": 3.8614,
+      "step": 982000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.0429529608794375e-05,
+      "loss": 3.8612,
+      "step": 983000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.024736926677754e-05,
+      "loss": 3.8581,
+      "step": 984000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.0066951746339515e-05,
+      "loss": 3.8562,
+      "step": 985000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 9.887917337602925e-06,
+      "loss": 3.8566,
+      "step": 986000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 9.71044889515631e-06,
+      "loss": 3.8534,
+      "step": 987000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 9.534722998420087e-06,
+      "loss": 3.8577,
+      "step": 988000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 9.360389291505156e-06,
+      "loss": 3.8599,
+      "step": 989000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 9.187625896164997e-06,
+      "loss": 3.8554,
+      "step": 990000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 9.016434235463455e-06,
+      "loss": 3.8554,
+      "step": 991000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.846984551782144e-06,
+      "loss": 3.8531,
+      "step": 992000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.678939002516817e-06,
+      "loss": 3.8537,
+      "step": 993000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.512635059971796e-06,
+      "loss": 3.8556,
+      "step": 994000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.34774115340684e-06,
+      "loss": 3.8552,
+      "step": 995000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.184588425936723e-06,
+      "loss": 3.8518,
+      "step": 996000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.022851589599123e-06,
+      "loss": 3.8519,
+      "step": 997000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 7.862855447419604e-06,
+      "loss": 3.8497,
+      "step": 998000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 7.70428100492051e-06,
+      "loss": 3.8534,
+      "step": 999000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 7.54729051547387e-06,
+      "loss": 3.8516,
+      "step": 1000000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.392039885206847e-06,
+      "loss": 3.8476,
+      "step": 1001000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.238219581070471e-06,
+      "loss": 3.8473,
+      "step": 1002000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 7.085987068966549e-06,
+      "loss": 3.8497,
+      "step": 1003000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 6.935343602844757e-06,
+      "loss": 3.8455,
+      "step": 1004000
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 6.786438681986962e-06,
+      "loss": 3.8503,
+      "step": 1005000
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 6.638975425188365e-06,
+      "loss": 3.8475,
+      "step": 1006000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 6.493249970997628e-06,
+      "loss": 3.8456,
+      "step": 1007000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 6.349115258944571e-06,
+      "loss": 3.8407,
+      "step": 1008000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 6.2064305863833495e-06,
+      "loss": 3.8426,
+      "step": 1009000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 6.065342204771441e-06,
+      "loss": 3.8456,
+      "step": 1010000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 5.92585127626355e-06,
+      "loss": 3.8446,
+      "step": 1011000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 5.7880960433015715e-06,
+      "loss": 3.8439,
+      "step": 1012000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 5.651801854522143e-06,
+      "loss": 3.8432,
+      "step": 1013000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 5.517108525207015e-06,
+      "loss": 3.8411,
+      "step": 1014000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 5.384017164834387e-06,
+      "loss": 3.8405,
+      "step": 1015000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 5.25265955688945e-06,
+      "loss": 3.8406,
+      "step": 1016000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 5.122773805360459e-06,
+      "loss": 3.8425,
+      "step": 1017000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 4.9946207493118515e-06,
+      "loss": 3.8429,
+      "step": 1018000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 4.867944881850673e-06,
+      "loss": 3.84,
+      "step": 1019000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 4.743000595890457e-06,
+      "loss": 3.8434,
+      "step": 1020000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 4.619538782067134e-06,
+      "loss": 3.8404,
+      "step": 1021000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 4.497686330529982e-06,
+      "loss": 3.8382,
+      "step": 1022000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 4.377444244986006e-06,
+      "loss": 3.8392,
+      "step": 1023000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 4.2589313414077795e-06,
+      "loss": 3.8395,
+      "step": 1024000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 4.1420275474132856e-06,
+      "loss": 3.8404,
+      "step": 1025000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 4.02661922185521e-06,
+      "loss": 3.841,
+      "step": 1026000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 3.912938129952815e-06,
+      "loss": 3.8376,
+      "step": 1027000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 3.8007576184877935e-06,
+      "loss": 3.8383,
+      "step": 1028000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 3.6901932136656604e-06,
+      "loss": 3.837,
+      "step": 1029000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 3.5812458262129755e-06,
+      "loss": 3.8348,
+      "step": 1030000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 3.4740228745658187e-06,
+      "loss": 3.8393,
+      "step": 1031000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 3.368310581510614e-06,
+      "loss": 3.8335,
+      "step": 1032000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 3.2643212405075284e-06,
+      "loss": 3.833,
+      "step": 1033000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 3.161949185382773e-06,
+      "loss": 3.8357,
+      "step": 1034000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 3.0610952138760753e-06,
+      "loss": 3.8334,
+      "step": 1035000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.9618634416622936e-06,
+      "loss": 3.8355,
+      "step": 1036000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.864351483910399e-06,
+      "loss": 3.8369,
+      "step": 1037000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.768364924832545e-06,
+      "loss": 3.8369,
+      "step": 1038000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.6740029762871932e-06,
+      "loss": 3.8313,
+      "step": 1039000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.581358339964313e-06,
+      "loss": 3.8306,
+      "step": 1040000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.4902463043641854e-06,
+      "loss": 3.8327,
+      "step": 1041000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.4007611701787116e-06,
+      "loss": 3.834,
+      "step": 1042000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.3129036745030752e-06,
+      "loss": 3.8316,
+      "step": 1043000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.226759956554547e-06,
+      "loss": 3.8326,
+      "step": 1044000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.1421582661275585e-06,
+      "loss": 3.8354,
+      "step": 1045000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.0592685019640958e-06,
+      "loss": 3.8307,
+      "step": 1046000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.977925401473013e-06,
+      "loss": 3.8299,
+      "step": 1047000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.8982923194333036e-06,
+      "loss": 3.8297,
+      "step": 1048000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.8202104863079827e-06,
+      "loss": 3.8298,
+      "step": 1049000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.7438367075362172e-06,
+      "loss": 3.8303,
+      "step": 1050000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.6690927139422218e-06,
+      "loss": 3.8317,
+      "step": 1051000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.5959067502205883e-06,
+      "loss": 3.8296,
+      "step": 1052000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.5243550546499618e-06,
+      "loss": 3.8309,
+      "step": 1053000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.4544382166065795e-06,
+      "loss": 3.8339,
+      "step": 1054000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.3861568120002276e-06,
+      "loss": 3.831,
+      "step": 1055000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.3195114032695576e-06,
+      "loss": 3.8265,
+      "step": 1056000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.2545667306077758e-06,
+      "loss": 3.8309,
+      "step": 1057000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1911933096932392e-06,
+      "loss": 3.8254,
+      "step": 1058000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.1294574905821087e-06,
+      "loss": 3.8288,
+      "step": 1059000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0694190611034273e-06,
+      "loss": 3.8261,
+      "step": 1060000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.0109583188243843e-06,
+      "loss": 3.8253,
+      "step": 1061000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 9.541366629567838e-07,
+      "loss": 3.8274,
+      "step": 1062000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 8.989545615444961e-07,
+      "loss": 3.8264,
+      "step": 1063000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 8.454651918863299e-07,
+      "loss": 3.8278,
+      "step": 1064000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.935619088263124e-07,
+      "loss": 3.8267,
+      "step": 1065000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.433489455357823e-07,
+      "loss": 3.8298,
+      "step": 1066000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 6.94726189238426e-07,
+      "loss": 3.8305,
+      "step": 1067000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 6.477912850886725e-07,
+      "loss": 3.8305,
+      "step": 1068000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 6.024951762708009e-07,
+      "loss": 3.8248,
+      "step": 1069000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 5.588382351461308e-07,
+      "loss": 3.826,
+      "step": 1070000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 5.16779582648863e-07,
+      "loss": 3.8261,
+      "step": 1071000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 4.76364087147263e-07,
+      "loss": 3.8238,
+      "step": 1072000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 4.375920815465229e-07,
+      "loss": 3.8293,
+      "step": 1073000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 4.004638852143083e-07,
+      "loss": 3.8244,
+      "step": 1074000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 3.6497980397816043e-07,
+      "loss": 3.8275,
+      "step": 1075000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 3.3117314832133985e-07,
+      "loss": 3.8278,
+      "step": 1076000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.989765157657809e-07,
+      "loss": 3.824,
+      "step": 1077000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.684545642082537e-07,
+      "loss": 3.8268,
+      "step": 1078000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 2.395464400940739e-07,
+      "loss": 3.8244,
+      "step": 1079000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.1228375656396903e-07,
+      "loss": 3.8236,
+      "step": 1080000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.8666673818257262e-07,
+      "loss": 3.8255,
+      "step": 1081000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.6271874491924355e-07,
+      "loss": 3.828,
+      "step": 1082000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.4041353457650008e-07,
+      "loss": 3.8259,
+      "step": 1083000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.1973143077612658e-07,
+      "loss": 3.8265,
+      "step": 1084000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.0069575448430346e-07,
+      "loss": 3.826,
+      "step": 1085000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 8.330666249920515e-08,
+      "loss": 3.8267,
+      "step": 1086000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 6.759413926236135e-08,
+      "loss": 3.8269,
+      "step": 1087000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 5.349533819716257e-08,
+      "loss": 3.8255,
+      "step": 1088000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 4.1043510231775216e-08,
+      "loss": 3.8275,
+      "step": 1089000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 3.0248739940019756e-08,
+      "loss": 3.8277,
+      "step": 1090000
+    }
+  ],
+  "max_steps": 1095620,
+  "num_train_epochs": 5,
+  "total_flos": 7.86932880566174e+19,
+  "trial_name": null,
+  "trial_params": null
+}