diff --git "a/checkpoint-100000/trainer_state.json" "b/checkpoint-100000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-100000/trainer_state.json"
@@ -0,0 +1,6716 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.3572704537334763,
+  "global_step": 100000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 1.0412,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 0.835,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 0.7822,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 0.7718,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 0.7707,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 0.7697,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1e-05,
+      "loss": 0.769,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 0.7682,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 0.7674,
+      "step": 900
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 0.767,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "eval_runtime": 45.7675,
+      "eval_samples_per_second": 235.975,
+      "eval_steps_per_second": 7.385,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 0.7665,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 0.7662,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.9e-05,
+      "loss": 0.7661,
+      "step": 1300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.2e-05,
+      "loss": 0.766,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 0.7659,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 0.7656,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.1e-05,
+      "loss": 0.7655,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 0.7655,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 0.7653,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.7655,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "eval_runtime": 45.5917,
+      "eval_samples_per_second": 236.885,
+      "eval_steps_per_second": 7.414,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 0.7651,
+      "step": 2100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 0.7653,
+      "step": 2200
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 6.9e-05,
+      "loss": 0.7654,
+      "step": 2300
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 0.765,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.5e-05,
+      "loss": 0.7649,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 7.8e-05,
+      "loss": 0.7648,
+      "step": 2600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 8.1e-05,
+      "loss": 0.7647,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.4e-05,
+      "loss": 0.7645,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 0.7645,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.7644,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "eval_runtime": 45.7281,
+      "eval_samples_per_second": 236.179,
+      "eval_steps_per_second": 7.392,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.3e-05,
+      "loss": 0.7641,
+      "step": 3100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 0.764,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.9e-05,
+      "loss": 0.7638,
+      "step": 3300
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000102,
+      "loss": 0.763,
+      "step": 3400
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 0.7665,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 0.7669,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 0.7653,
+      "step": 3700
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 0.7535,
+      "step": 3800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000117,
+      "loss": 0.7218,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.6956,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "eval_runtime": 45.9121,
+      "eval_samples_per_second": 235.232,
+      "eval_steps_per_second": 7.362,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 0.6758,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 0.6557,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.000129,
+      "loss": 0.6402,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 0.6302,
+      "step": 4400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000135,
+      "loss": 0.623,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000138,
+      "loss": 0.6169,
+      "step": 4600
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 0.6121,
+      "step": 4700
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 0.607,
+      "step": 4800
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000147,
+      "loss": 0.6039,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00015,
+      "loss": 0.6012,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "eval_runtime": 46.0979,
+      "eval_samples_per_second": 234.284,
+      "eval_steps_per_second": 7.332,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001499996172456075,
+      "loss": 0.5981,
+      "step": 5100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00014999846898661572,
+      "loss": 0.5954,
+      "step": 5200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00014999655523558183,
+      "loss": 0.5935,
+      "step": 5300
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00014999387601343436,
+      "loss": 0.5911,
+      "step": 5400
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00014999043134947282,
+      "loss": 0.5895,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00014998622128136748,
+      "loss": 0.5877,
+      "step": 5600
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000149981245855159,
+      "loss": 0.5866,
+      "step": 5700
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00014997550512525784,
+      "loss": 0.5845,
+      "step": 5800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001499689991544437,
+      "loss": 0.5784,
+      "step": 5900
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00014996172801386482,
+      "loss": 0.5684,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14,
+      "eval_runtime": 46.0154,
+      "eval_samples_per_second": 234.704,
+      "eval_steps_per_second": 7.345,
+      "step": 6000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00014995369178303722,
+      "loss": 0.5642,
+      "step": 6100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001499448905498439,
+      "loss": 0.5625,
+      "step": 6200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00014993532441053364,
+      "loss": 0.5601,
+      "step": 6300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001499249934697203,
+      "loss": 0.5581,
+      "step": 6400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001499138978403813,
+      "loss": 0.554,
+      "step": 6500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00014990203764385677,
+      "loss": 0.5462,
+      "step": 6600
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00014988941300984784,
+      "loss": 0.5284,
+      "step": 6700
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0001498760240764155,
+      "loss": 0.5032,
+      "step": 6800
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000149861870989979,
+      "loss": 0.4751,
+      "step": 6900
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001498469539053142,
+      "loss": 0.4574,
+      "step": 7000
+    },
+    {
+      "epoch": 0.17,
+      "eval_runtime": 45.9402,
+      "eval_samples_per_second": 235.088,
+      "eval_steps_per_second": 7.357,
+      "step": 7000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00014983127298555198,
+      "loss": 0.4453,
+      "step": 7100
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00014981482840217632,
+      "loss": 0.437,
+      "step": 7200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00014979762033502262,
+      "loss": 0.4306,
+      "step": 7300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00014977964897227547,
+      "loss": 0.4254,
+      "step": 7400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00014976091451046687,
+      "loss": 0.4204,
+      "step": 7500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00014974141715447386,
+      "loss": 0.4178,
+      "step": 7600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00014972115711751644,
+      "loss": 0.4135,
+      "step": 7700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00014970013462115505,
+      "loss": 0.4099,
+      "step": 7800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00014967834989528843,
+      "loss": 0.4077,
+      "step": 7900
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00014965580317815078,
+      "loss": 0.405,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "eval_runtime": 45.7648,
+      "eval_samples_per_second": 235.989,
+      "eval_steps_per_second": 7.386,
+      "step": 8000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00014963249471630944,
+      "loss": 0.4017,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000149608424764662,
+      "loss": 0.4006,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001495835935864336,
+      "loss": 0.3977,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00014955800145317397,
+      "loss": 0.3964,
+      "step": 8400
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00014953164864475466,
+      "loss": 0.3949,
+      "step": 8500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0001495045354493657,
+      "loss": 0.3961,
+      "step": 8600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00014947666216351272,
+      "loss": 0.398,
+      "step": 8700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00014944802909201344,
+      "loss": 0.3924,
+      "step": 8800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00014941863654799456,
+      "loss": 0.3938,
+      "step": 8900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00014938848485288825,
+      "loss": 0.3885,
+      "step": 9000
+    },
+    {
+      "epoch": 0.21,
+      "eval_runtime": 45.9868,
+      "eval_samples_per_second": 234.85,
+      "eval_steps_per_second": 7.35,
+      "step": 9000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001493575743364286,
+      "loss": 0.391,
+      "step": 9100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00014932590533664808,
+      "loss": 0.3884,
+      "step": 9200
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001492934781998738,
+      "loss": 0.3856,
+      "step": 9300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001492602932807237,
+      "loss": 0.3843,
+      "step": 9400
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00014922635094210277,
+      "loss": 0.3848,
+      "step": 9500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000149191651555199,
+      "loss": 0.3795,
+      "step": 9600
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001491561954994793,
+      "loss": 0.3735,
+      "step": 9700
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00014911998316268537,
+      "loss": 0.3658,
+      "step": 9800
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00014908301494082963,
+      "loss": 0.362,
+      "step": 9900
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00014904529123819054,
+      "loss": 0.3595,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "eval_runtime": 46.3224,
+      "eval_samples_per_second": 233.148,
+      "eval_steps_per_second": 7.297,
+      "step": 10000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00014900681246730852,
+      "loss": 0.3585,
+      "step": 10100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00014896757904898125,
+      "loss": 0.3578,
+      "step": 10200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00014892759141225904,
+      "loss": 0.3568,
+      "step": 10300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00014888684999444035,
+      "loss": 0.355,
+      "step": 10400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00014884535524106675,
+      "loss": 0.3537,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00014880310760591824,
+      "loss": 0.3523,
+      "step": 10600
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001487601075510082,
+      "loss": 0.3524,
+      "step": 10700
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001487163555465783,
+      "loss": 0.3515,
+      "step": 10800
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001486718520710935,
+      "loss": 0.3508,
+      "step": 10900
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00014862659761123663,
+      "loss": 0.3493,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "eval_runtime": 46.1625,
+      "eval_samples_per_second": 233.956,
+      "eval_steps_per_second": 7.322,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00014858059266190327,
+      "loss": 0.3472,
+      "step": 11100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014853383772619612,
+      "loss": 0.3463,
+      "step": 11200
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014848633331541967,
+      "loss": 0.3363,
+      "step": 11300
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001484380799490746,
+      "loss": 0.3265,
+      "step": 11400
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014838907815485194,
+      "loss": 0.3235,
+      "step": 11500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014833932846862748,
+      "loss": 0.3218,
+      "step": 11600
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014828883143445582,
+      "loss": 0.3203,
+      "step": 11700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001482375876045644,
+      "loss": 0.3204,
+      "step": 11800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001481855975393476,
+      "loss": 0.3184,
+      "step": 11900
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001481328618073604,
+      "loss": 0.318,
+      "step": 12000
+    },
+    {
+      "epoch": 0.29,
+      "eval_runtime": 46.1354,
+      "eval_samples_per_second": 234.094,
+      "eval_steps_per_second": 7.326,
+      "step": 12000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001480793809853123,
+      "loss": 0.3163,
+      "step": 12100
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014802515565806107,
+      "loss": 0.3155,
+      "step": 12200
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014797018641860612,
+      "loss": 0.314,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001479144738680823,
+      "loss": 0.3136,
+      "step": 12400
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014785801861575312,
+      "loss": 0.3117,
+      "step": 12500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014780082127900416,
+      "loss": 0.3086,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014774288248333635,
+      "loss": 0.3074,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014768420286235908,
+      "loss": 0.3074,
+      "step": 12800
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014762478305778328,
+      "loss": 0.3064,
+      "step": 12900
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001475646237194144,
+      "loss": 0.3057,
+      "step": 13000
+    },
+    {
+      "epoch": 0.31,
+      "eval_runtime": 46.1242,
+      "eval_samples_per_second": 234.15,
+      "eval_steps_per_second": 7.328,
+      "step": 13000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014750372550514533,
+      "loss": 0.3048,
+      "step": 13100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001474420890809492,
+      "loss": 0.3037,
+      "step": 13200
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014737971512087202,
+      "loss": 0.3029,
+      "step": 13300
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014731660430702552,
+      "loss": 0.3024,
+      "step": 13400
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014725275732957937,
+      "loss": 0.3011,
+      "step": 13500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00014718817488675387,
+      "loss": 0.3006,
+      "step": 13600
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00014712285768481235,
+      "loss": 0.3009,
+      "step": 13700
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00014705680643805323,
+      "loss": 0.2991,
+      "step": 13800
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00014699002186880232,
+      "loss": 0.2991,
+      "step": 13900
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00014692250470740503,
+      "loss": 0.2979,
+      "step": 14000
+    },
+    {
+      "epoch": 0.33,
+      "eval_runtime": 46.2531,
+      "eval_samples_per_second": 233.498,
+      "eval_steps_per_second": 7.308,
+      "step": 14000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00014685425569221819,
+      "loss": 0.2975,
+      "step": 14100
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00014678527556960207,
+      "loss": 0.2955,
+      "step": 14200
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001467155650939123,
+      "loss": 0.295,
+      "step": 14300
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00014664512502749141,
+      "loss": 0.2941,
+      "step": 14400
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014657395614066075,
+      "loss": 0.2931,
+      "step": 14500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001465020592117118,
+      "loss": 0.2921,
+      "step": 14600
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001464294350268979,
+      "loss": 0.2918,
+      "step": 14700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014635608438042546,
+      "loss": 0.2907,
+      "step": 14800
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014628200807444543,
+      "loss": 0.2899,
+      "step": 14900
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001462072069190444,
+      "loss": 0.2898,
+      "step": 15000
+    },
+    {
+      "epoch": 0.36,
+      "eval_runtime": 46.2774,
+      "eval_samples_per_second": 233.375,
+      "eval_steps_per_second": 7.304,
+      "step": 15000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014613168173223585,
+      "loss": 0.2885,
+      "step": 15100
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014605543333995113,
+      "loss": 0.288,
+      "step": 15200
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00014597846257603038,
+      "loss": 0.2875,
+      "step": 15300
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001459007702822136,
+      "loss": 0.2876,
+      "step": 15400
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014582235730813128,
+      "loss": 0.2862,
+      "step": 15500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014574322451129507,
+      "loss": 0.2849,
+      "step": 15600
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014566337275708863,
+      "loss": 0.2852,
+      "step": 15700
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0001455828029187579,
+      "loss": 0.2833,
+      "step": 15800
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014550151587740178,
+      "loss": 0.2836,
+      "step": 15900
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014541951252196225,
+      "loss": 0.2817,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38,
+      "eval_runtime": 46.1169,
+      "eval_samples_per_second": 234.187,
+      "eval_steps_per_second": 7.329,
+      "step": 16000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014533679374921493,
+      "loss": 0.2824,
+      "step": 16100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014525336046375905,
+      "loss": 0.2817,
+      "step": 16200
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014516921357800766,
+      "loss": 0.2812,
+      "step": 16300
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014508435401217759,
+      "loss": 0.2812,
+      "step": 16400
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014499878269427948,
+      "loss": 0.2795,
+      "step": 16500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00014491250056010758,
+      "loss": 0.2788,
+      "step": 16600
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00014482550855322943,
+      "loss": 0.2775,
+      "step": 16700
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001447378076249757,
+      "loss": 0.2773,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00014464939873442973,
+      "loss": 0.2769,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00014456028284841693,
+      "loss": 0.2765,
+      "step": 17000
+    },
+    {
+      "epoch": 0.4,
+      "eval_runtime": 46.3516,
+      "eval_samples_per_second": 233.002,
+      "eval_steps_per_second": 7.292,
+      "step": 17000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00014447046094149437,
+      "loss": 0.2752,
+      "step": 17100
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00014437993399594003,
+      "loss": 0.2765,
+      "step": 17200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001442887030017421,
+      "loss": 0.2752,
+      "step": 17300
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00014419676895658807,
+      "loss": 0.2748,
+      "step": 17400
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.000144104132865854,
+      "loss": 0.2739,
+      "step": 17500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001440107957425933,
+      "loss": 0.2729,
+      "step": 17600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001439167586075258,
+      "loss": 0.2722,
+      "step": 17700
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001438220224890265,
+      "loss": 0.2725,
+      "step": 17800
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00014372658842311449,
+      "loss": 0.2726,
+      "step": 17900
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00014363045745344137,
+      "loss": 0.2715,
+      "step": 18000
+    },
+    {
+      "epoch": 0.43,
+      "eval_runtime": 46.2247,
+      "eval_samples_per_second": 233.641,
+      "eval_steps_per_second": 7.312,
+      "step": 18000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00014353363063128005,
+      "loss": 0.2705,
+      "step": 18100
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001434361090155131,
+      "loss": 0.2706,
+      "step": 18200
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00014333789367262136,
+      "loss": 0.2701,
+      "step": 18300
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00014323898567667202,
+      "loss": 0.2693,
+      "step": 18400
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00014313938610930712,
+      "loss": 0.2693,
+      "step": 18500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00014303909605973154,
+      "loss": 0.2691,
+      "step": 18600
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001429381166247012,
+      "loss": 0.2681,
+      "step": 18700
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00014283644890851103,
+      "loss": 0.2672,
+      "step": 18800
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00014273409402298291,
+      "loss": 0.2671,
+      "step": 18900
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00014263105308745343,
+      "loss": 0.2676,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45,
+      "eval_runtime": 46.3331,
+      "eval_samples_per_second": 233.095,
+      "eval_steps_per_second": 7.295,
+      "step": 19000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00014252732722876176,
+      "loss": 0.2654,
+      "step": 19100
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001424229175812373,
+      "loss": 0.2649,
+      "step": 19200
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00014231782528668717,
+      "loss": 0.2647,
+      "step": 19300
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00014221205149438394,
+      "loss": 0.2649,
+      "step": 19400
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001421055973610528,
+      "loss": 0.264,
+      "step": 19500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00014199846405085913,
+      "loss": 0.2647,
+      "step": 19600
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00014189065273539564,
+      "loss": 0.2635,
+      "step": 19700
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00014178216459366958,
+      "loss": 0.2623,
+      "step": 19800
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00014167300081208988,
+      "loss": 0.2627,
+      "step": 19900
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00014156316258445421,
+      "loss": 0.2932,
+      "step": 20000
+    },
+    {
+      "epoch": 0.48,
+      "eval_runtime": 46.169,
+      "eval_samples_per_second": 233.923,
+      "eval_steps_per_second": 7.321,
+      "step": 20000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00014145265111193583,
+      "loss": 0.2645,
+      "step": 20100
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00014134146760307043,
+      "loss": 0.2625,
+      "step": 20200
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00014122961327374313,
+      "loss": 0.2615,
+      "step": 20300
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001411170893471749,
+      "loss": 0.2605,
+      "step": 20400
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00014100389705390938,
+      "loss": 0.26,
+      "step": 20500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001408900376317994,
+      "loss": 0.2583,
+      "step": 20600
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001407755123259933,
+      "loss": 0.258,
+      "step": 20700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014066032238892152,
+      "loss": 0.2569,
+      "step": 20800
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014054446908028272,
+      "loss": 0.2568,
+      "step": 20900
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014042795366703018,
+      "loss": 0.2563,
+      "step": 21000
+    },
+    {
+      "epoch": 0.5,
+      "eval_runtime": 46.2726,
+      "eval_samples_per_second": 233.4,
+      "eval_steps_per_second": 7.305,
+      "step": 21000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001403107774233577,
+      "loss": 0.256,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014019294163068597,
+      "loss": 0.2548,
+      "step": 21200
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014007444757764835,
+      "loss": 0.2543,
+      "step": 21300
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001399552965600768,
+      "loss": 0.2537,
+      "step": 21400
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001398354898809877,
+      "loss": 0.2531,
+      "step": 21500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001397150288505678,
+      "loss": 0.2531,
+      "step": 21600
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013959391478615959,
+      "loss": 0.2526,
+      "step": 21700
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013947214901224706,
+      "loss": 0.2522,
+      "step": 21800
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001393497328604412,
+      "loss": 0.2515,
+      "step": 21900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00013922666766946545,
+      "loss": 0.2513,
+      "step": 22000
+    },
+    {
+      "epoch": 0.52,
+      "eval_runtime": 46.224,
+      "eval_samples_per_second": 233.645,
+      "eval_steps_per_second": 7.312,
+      "step": 22000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013910295478514106,
+      "loss": 0.2504,
+      "step": 22100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001389785955603722,
+      "loss": 0.2503,
+      "step": 22200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00013885359135513154,
+      "loss": 0.2501,
+      "step": 22300
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000138727943536445,
+      "loss": 0.2488,
+      "step": 22400
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013860165347837698,
+      "loss": 0.2492,
+      "step": 22500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013847472256201535,
+      "loss": 0.2483,
+      "step": 22600
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013834715217545625,
+      "loss": 0.248,
+      "step": 22700
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000138218943713789,
+      "loss": 0.2479,
+      "step": 22800
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001380900985790808,
+      "loss": 0.2485,
+      "step": 22900
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013796061818036138,
+      "loss": 0.2467,
+      "step": 23000
+    },
+    {
+      "epoch": 0.55,
+      "eval_runtime": 46.1546,
+      "eval_samples_per_second": 233.996,
+      "eval_steps_per_second": 7.323,
+      "step": 23000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013783050393360768,
+      "loss": 0.2468,
+      "step": 23100
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001376997572617282,
+      "loss": 0.2463,
+      "step": 23200
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013756837959454766,
+      "loss": 0.2456,
+      "step": 23300
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001374363723687911,
+      "loss": 0.2459,
+      "step": 23400
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013730373702806846,
+      "loss": 0.2447,
+      "step": 23500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013717047502285855,
+      "loss": 0.245,
+      "step": 23600
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001370365878104933,
+      "loss": 0.2446,
+      "step": 23700
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00013690207685514185,
+      "loss": 0.2442,
+      "step": 23800
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001367669436277944,
+      "loss": 0.2439,
+      "step": 23900
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001366311896062463,
+      "loss": 0.2438,
+      "step": 24000
+    },
+    {
+      "epoch": 0.57,
+      "eval_runtime": 46.5558,
+      "eval_samples_per_second": 231.98,
+      "eval_steps_per_second": 7.26,
+      "step": 24000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00013649481627508181,
+      "loss": 0.2436,
+      "step": 24100
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001363578251256578,
+      "loss": 0.2429,
+      "step": 24200
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00013622021765608754,
+      "loss": 0.2424,
+      "step": 24300
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00013608199537122425,
+      "loss": 0.242,
+      "step": 24400
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001359431597826447,
+      "loss": 0.2422,
+      "step": 24500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001358037124086327,
+      "loss": 0.2418,
+      "step": 24600
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00013566365477416233,
+      "loss": 0.2407,
+      "step": 24700
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00013552298841088144,
+      "loss": 0.2416,
+      "step": 24800
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00013538171485709486,
+      "loss": 0.2411,
+      "step": 24900
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00013523983565774753,
+      "loss": 0.2401,
+      "step": 25000
+    },
+    {
+      "epoch": 0.6,
+      "eval_runtime": 46.0773,
+      "eval_samples_per_second": 234.389,
+      "eval_steps_per_second": 7.336,
+      "step": 25000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00013509735236440766,
+      "loss": 0.2401,
+      "step": 25100
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00013495426653524972,
+      "loss": 0.2402,
+      "step": 25200
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00013481057973503742,
+      "loss": 0.24,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00013466629353510651,
+      "loss": 0.239,
+      "step": 25400
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013452140951334787,
+      "loss": 0.239,
+      "step": 25500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013437592925418985,
+      "loss": 0.2388,
+      "step": 25600
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013422985434858133,
+      "loss": 0.238,
+      "step": 25700
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013408318639397405,
+      "loss": 0.2387,
+      "step": 25800
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013393592699430525,
+      "loss": 0.2372,
+      "step": 25900
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013378807775998012,
+      "loss": 0.2377,
+      "step": 26000
+    },
+    {
+      "epoch": 0.62,
+      "eval_runtime": 46.2501,
+      "eval_samples_per_second": 233.513,
+      "eval_steps_per_second": 7.308,
+      "step": 26000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013363964030785422,
+      "loss": 0.2373,
+      "step": 26100
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013349061626121578,
+      "loss": 0.238,
+      "step": 26200
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013334100724976783,
+      "loss": 0.2367,
+      "step": 26300
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001331908149096106,
+      "loss": 0.2367,
+      "step": 26400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013304004088322342,
+      "loss": 0.2356,
+      "step": 26500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013288868681944692,
+      "loss": 0.2365,
+      "step": 26600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00013273675437346487,
+      "loss": 0.236,
+      "step": 26700
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00013258424520678618,
+      "loss": 0.2356,
+      "step": 26800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00013243116098722663,
+      "loss": 0.2363,
+      "step": 26900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00013227750338889077,
+      "loss": 0.2345,
+      "step": 27000
+    },
+    {
+      "epoch": 0.64,
+      "eval_runtime": 46.2738,
+      "eval_samples_per_second": 233.394,
+      "eval_steps_per_second": 7.304,
+      "step": 27000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00013212327409215343,
+      "loss": 0.2351,
+      "step": 27100
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001319684747836415,
+      "loss": 0.2351,
+      "step": 27200
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001318131071562154,
+      "loss": 0.2342,
+      "step": 27300
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00013165717290895067,
+      "loss": 0.2338,
+      "step": 27400
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001315006737471192,
+      "loss": 0.234,
+      "step": 27500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001313436113821708,
+      "loss": 0.233,
+      "step": 27600
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00013118598753171425,
+      "loss": 0.2331,
+      "step": 27700
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001310278039194988,
+      "loss": 0.2329,
+      "step": 27800
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00013086906227539506,
+      "loss": 0.2332,
+      "step": 27900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00013070976433537623,
+      "loss": 0.2338,
+      "step": 28000
+    },
+    {
+      "epoch": 0.67,
+      "eval_runtime": 46.2625,
+      "eval_samples_per_second": 233.45,
+      "eval_steps_per_second": 7.306,
+      "step": 28000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00013054991184149905,
+      "loss": 0.2325,
+      "step": 28100
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00013038950654188476,
+      "loss": 0.2312,
+      "step": 28200
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00013022855019070005,
+      "loss": 0.2323,
+      "step": 28300
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001300670445481378,
+      "loss": 0.2319,
+      "step": 28400
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001299049913803978,
+      "loss": 0.2324,
+      "step": 28500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00012974239245966754,
+      "loss": 0.2313,
+      "step": 28600
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001295792495641028,
+      "loss": 0.2318,
+      "step": 28700
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00012941556447780813,
+      "loss": 0.2309,
+      "step": 28800
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001292513389908174,
+      "loss": 0.231,
+      "step": 28900
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001290865748990742,
+      "loss": 0.2298,
+      "step": 29000
+    },
+    {
+      "epoch": 0.69,
+      "eval_runtime": 46.1555,
+      "eval_samples_per_second": 233.992,
+      "eval_steps_per_second": 7.323,
+      "step": 29000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00012892127400441228,
+      "loss": 0.2302,
+      "step": 29100
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00012875543811453576,
+      "loss": 0.2305,
+      "step": 29200
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001285890690429993,
+      "loss": 0.2293,
+      "step": 29300
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00012842216860918846,
+      "loss": 0.2298,
+      "step": 29400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001282547386382996,
+      "loss": 0.2296,
+      "step": 29500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001280867809613201,
+      "loss": 0.2291,
+      "step": 29600
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001279182974150082,
+      "loss": 0.2279,
+      "step": 29700
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00012774928984187297,
+      "loss": 0.2278,
+      "step": 29800
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00012757976009015413,
+      "loss": 0.228,
+      "step": 29900
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001274097100138019,
+      "loss": 0.2282,
+      "step": 30000
+    },
+    {
+      "epoch": 0.71,
+      "eval_runtime": 46.6895,
+      "eval_samples_per_second": 231.315,
+      "eval_steps_per_second": 7.239,
+      "step": 30000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00012723914147245663,
+      "loss": 0.2276,
+      "step": 30100
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00012706805633142863,
+      "loss": 0.2276,
+      "step": 30200
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00012689645646167755,
+      "loss": 0.2281,
+      "step": 30300
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00012672434373979207,
+      "loss": 0.2265,
+      "step": 30400
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00012655172004796936,
+      "loss": 0.2286,
+      "step": 30500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00012637858727399448,
+      "loss": 0.227,
+      "step": 30600
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00012620494731121966,
+      "loss": 0.2267,
+      "step": 30700
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00012603080205854372,
+      "loss": 0.2266,
+      "step": 30800
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00012585615342039126,
+      "loss": 0.2258,
+      "step": 30900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001256810033066918,
+      "loss": 0.226,
+      "step": 31000
+    },
+    {
+      "epoch": 0.74,
+      "eval_runtime": 47.0689,
+      "eval_samples_per_second": 229.451,
+      "eval_steps_per_second": 7.181,
+      "step": 31000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001255053536328589,
+      "loss": 0.2257,
+      "step": 31100
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001253292063197693,
+      "loss": 0.2256,
+      "step": 31200
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001251525632937418,
+      "loss": 0.2257,
+      "step": 31300
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00012497542648651615,
+      "loss": 0.2248,
+      "step": 31400
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00012479779783523216,
+      "loss": 0.225,
+      "step": 31500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00012461967928240828,
+      "loss": 0.2246,
+      "step": 31600
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00012444107277592047,
+      "loss": 0.2247,
+      "step": 31700
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001242619802689809,
+      "loss": 0.2246,
+      "step": 31800
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00012408240372011647,
+      "loss": 0.2238,
+      "step": 31900
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001239023450931476,
+      "loss": 0.2243,
+      "step": 32000
+    },
+    {
+      "epoch": 0.76,
+      "eval_runtime": 47.1954,
+      "eval_samples_per_second": 228.836,
+      "eval_steps_per_second": 7.162,
+      "step": 32000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00012372180635716656,
+      "loss": 0.2235,
+      "step": 32100
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00012354078948651604,
+      "loss": 0.2239,
+      "step": 32200
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00012335929646076758,
+      "loss": 0.2231,
+      "step": 32300
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00012317732926469976,
+      "loss": 0.2225,
+      "step": 32400
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00012299488988827675,
+      "loss": 0.2233,
+      "step": 32500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001228119803266263,
+      "loss": 0.223,
+      "step": 32600
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001226286025800181,
+      "loss": 0.2229,
+      "step": 32700
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00012244475865384177,
+      "loss": 0.222,
+      "step": 32800
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00012226045055858505,
+      "loss": 0.2217,
+      "step": 32900
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00012207568030981174,
+      "loss": 0.2222,
+      "step": 33000
+    },
+    {
+      "epoch": 0.79,
+      "eval_runtime": 47.0101,
+      "eval_samples_per_second": 229.738,
+      "eval_steps_per_second": 7.19,
+      "step": 33000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00012189044992813972,
+      "loss": 0.2213,
+      "step": 33100
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001217047614392187,
+      "loss": 0.2206,
+      "step": 33200
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00012151861687370828,
+      "loss": 0.2221,
+      "step": 33300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00012133201826725558,
+      "loss": 0.2209,
+      "step": 33400
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001211449676604731,
+      "loss": 0.2211,
+      "step": 33500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00012095746709891632,
+      "loss": 0.2205,
+      "step": 33600
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00012076951863306127,
+      "loss": 0.2203,
+      "step": 33700
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001205811243182823,
+      "loss": 0.22,
+      "step": 33800
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00012039228621482949,
+      "loss": 0.2192,
+      "step": 33900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00012020300638780604,
+      "loss": 0.219,
+      "step": 34000
+    },
+    {
+      "epoch": 0.81,
+      "eval_runtime": 47.0946,
+      "eval_samples_per_second": 229.325,
+      "eval_steps_per_second": 7.177,
+      "step": 34000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00012001328690714582,
+      "loss": 0.2194,
+      "step": 34100
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00011982312984759068,
+      "loss": 0.2194,
+      "step": 34200
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00011963253728866778,
+      "loss": 0.2189,
+      "step": 34300
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00011944151131466675,
+      "loss": 0.219,
+      "step": 34400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00011925005401461709,
+      "loss": 0.2184,
+      "step": 34500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00011905816748226513,
+      "loss": 0.2182,
+      "step": 34600
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00011886585381605125,
+      "loss": 0.2188,
+      "step": 34700
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00011867311511908693,
+      "loss": 0.2179,
+      "step": 34800
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00011847995349913162,
+      "loss": 0.218,
+      "step": 34900
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00011828637106856989,
+      "loss": 0.2173,
+      "step": 35000
+    },
+    {
+      "epoch": 0.83,
+      "eval_runtime": 46.7598,
+      "eval_samples_per_second": 230.968,
+      "eval_steps_per_second": 7.228,
+      "step": 35000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00011809236994438816,
+      "loss": 0.2171,
+      "step": 35100
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00011789795224815164,
+      "loss": 0.2175,
+      "step": 35200
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00011770312010598116,
+      "loss": 0.2167,
+      "step": 35300
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00011750787564852973,
+      "loss": 0.2167,
+      "step": 35400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00011731222101095955,
+      "loss": 0.2171,
+      "step": 35500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00011711615833291833,
+      "loss": 0.2161,
+      "step": 35600
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001169196897585161,
+      "loss": 0.2168,
+      "step": 35700
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00011672281743630175,
+      "loss": 0.2162,
+      "step": 35800
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001165255435192394,
+      "loss": 0.2152,
+      "step": 35900
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00011632787016468506,
+      "loss": 0.216,
+      "step": 36000
+    },
+    {
+      "epoch": 0.86,
+      "eval_runtime": 47.0992,
+      "eval_samples_per_second": 229.303,
+      "eval_steps_per_second": 7.176,
+      "step": 36000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001161297995343628,
+      "loss": 0.2157,
+      "step": 36100
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00011593133379434138,
+      "loss": 0.215,
+      "step": 36200
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00011573247511501028,
+      "loss": 0.2154,
+      "step": 36300
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00011553322567105619,
+      "loss": 0.2155,
+      "step": 36400
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00011533358764143905,
+      "loss": 0.2149,
+      "step": 36500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00011513356320936841,
+      "loss": 0.2144,
+      "step": 36600
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00011493315456227943,
+      "loss": 0.2147,
+      "step": 36700
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00011473236389180894,
+      "loss": 0.2145,
+      "step": 36800
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00011453119339377154,
+      "loss": 0.2146,
+      "step": 36900
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00011432964526813558,
+      "loss": 0.2145,
+      "step": 37000
+    },
+    {
+      "epoch": 0.88,
+      "eval_runtime": 46.8321,
+      "eval_samples_per_second": 230.611,
+      "eval_steps_per_second": 7.217,
+      "step": 37000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00011412772171899904,
+      "loss": 0.2132,
+      "step": 37100
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00011392542495456556,
+      "loss": 0.2133,
+      "step": 37200
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00011372275718712006,
+      "loss": 0.2125,
+      "step": 37300
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00011351972063300484,
+      "loss": 0.2135,
+      "step": 37400
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00011331631751259515,
+      "loss": 0.213,
+      "step": 37500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00011311255005027487,
+      "loss": 0.2132,
+      "step": 37600
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00011290842047441232,
+      "loss": 0.2125,
+      "step": 37700
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00011270393101733585,
+      "loss": 0.2122,
+      "step": 37800
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00011249908391530946,
+      "loss": 0.2113,
+      "step": 37900
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00011229388140850814,
+      "loss": 0.2119,
+      "step": 38000
+    },
+    {
+      "epoch": 0.91,
+      "eval_runtime": 46.8036,
+      "eval_samples_per_second": 230.751,
+      "eval_steps_per_second": 7.222,
+      "step": 38000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00011208832574099368,
+      "loss": 0.2113,
+      "step": 38100
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00011188241916068993,
+      "loss": 0.2111,
+      "step": 38200
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00011167616391935826,
+      "loss": 0.2111,
+      "step": 38300
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00011146956227257293,
+      "loss": 0.2119,
+      "step": 38400
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00011126261647969645,
+      "loss": 0.2115,
+      "step": 38500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00011105532880385487,
+      "loss": 0.2104,
+      "step": 38600
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00011084770151191299,
+      "loss": 0.2107,
+      "step": 38700
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00011063973687444962,
+      "loss": 0.2097,
+      "step": 38800
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00011043143716573272,
+      "loss": 0.2107,
+      "step": 38900
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00011022280466369448,
+      "loss": 0.2113,
+      "step": 39000
+    },
+    {
+      "epoch": 0.93,
+      "eval_runtime": 47.0898,
+      "eval_samples_per_second": 229.349,
+      "eval_steps_per_second": 7.178,
+      "step": 39000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00011001384164990662,
+      "loss": 0.2099,
+      "step": 39100
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00010980455040955506,
+      "loss": 0.21,
+      "step": 39200
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00010959493323141538,
+      "loss": 0.2091,
+      "step": 39300
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00010938499240782739,
+      "loss": 0.2098,
+      "step": 39400
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00010917473023467032,
+      "loss": 0.2096,
+      "step": 39500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00010896414901133761,
+      "loss": 0.2085,
+      "step": 39600
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00010875325104071177,
+      "loss": 0.2093,
+      "step": 39700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00010854203862913927,
+      "loss": 0.2084,
+      "step": 39800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00010833051408640509,
+      "loss": 0.2083,
+      "step": 39900
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00010811867972570786,
+      "loss": 0.2084,
+      "step": 40000
+    },
+    {
+      "epoch": 0.95,
+      "eval_runtime": 46.8854,
+      "eval_samples_per_second": 230.349,
+      "eval_steps_per_second": 7.209,
+      "step": 40000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00010790653786363416,
+      "loss": 0.2082,
+      "step": 40100
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00010769409082013337,
+      "loss": 0.2081,
+      "step": 40200
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00010748134091849238,
+      "loss": 0.2077,
+      "step": 40300
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00010726829048531,
+      "loss": 0.2078,
+      "step": 40400
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00010705494185047165,
+      "loss": 0.2077,
+      "step": 40500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001068412973471238,
+      "loss": 0.2073,
+      "step": 40600
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00010662735931164853,
+      "loss": 0.2076,
+      "step": 40700
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001064131300836379,
+      "loss": 0.2069,
+      "step": 40800
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001061986120058684,
+      "loss": 0.2067,
+      "step": 40900
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00010598380742427543,
+      "loss": 0.206,
+      "step": 41000
+    },
+    {
+      "epoch": 0.98,
+      "eval_runtime": 46.6481,
+      "eval_samples_per_second": 231.521,
+      "eval_steps_per_second": 7.246,
+      "step": 41000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00010576871868792746,
+      "loss": 0.206,
+      "step": 41100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001055533481490004,
+      "loss": 0.2058,
+      "step": 41200
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.000105337698162752,
+      "loss": 0.206,
+      "step": 41300
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00010512177108749594,
+      "loss": 0.2057,
+      "step": 41400
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00010490556928457616,
+      "loss": 0.2039,
+      "step": 41500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00010468909511834088,
+      "loss": 0.205,
+      "step": 41600
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00010447235095611692,
+      "loss": 0.2045,
+      "step": 41700
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00010425533916818376,
+      "loss": 0.2047,
+      "step": 41800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00010403806212774747,
+      "loss": 0.205,
+      "step": 41900
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.000103820522210915,
+      "loss": 0.2042,
+      "step": 42000
+    },
+    {
+      "epoch": 1.0,
+      "eval_runtime": 46.7967,
+      "eval_samples_per_second": 230.786,
+      "eval_steps_per_second": 7.223,
+      "step": 42000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00010360272179666802,
+      "loss": 0.204,
+      "step": 42100
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00010338466326683697,
+      "loss": 0.2037,
+      "step": 42200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00010316634900607497,
+      "loss": 0.2033,
+      "step": 42300
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00010294778140183182,
+      "loss": 0.2035,
+      "step": 42400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00010272896284432785,
+      "loss": 0.2037,
+      "step": 42500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00010250989572652766,
+      "loss": 0.2028,
+      "step": 42600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00010229058244411427,
+      "loss": 0.2019,
+      "step": 42700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00010207102539546251,
+      "loss": 0.2032,
+      "step": 42800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00010185122698161311,
+      "loss": 0.2026,
+      "step": 42900
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00010163118960624632,
+      "loss": 0.2024,
+      "step": 43000
+    },
+    {
+      "epoch": 1.02,
+      "eval_runtime": 46.9319,
+      "eval_samples_per_second": 230.121,
+      "eval_steps_per_second": 7.202,
+      "step": 43000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00010141091567565561,
+      "loss": 0.2028,
+      "step": 43100
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00010119040759872142,
+      "loss": 0.2018,
+      "step": 43200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00010096966778688472,
+      "loss": 0.2016,
+      "step": 43300
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00010074869865412074,
+      "loss": 0.2024,
+      "step": 43400
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00010052750261691254,
+      "loss": 0.2017,
+      "step": 43500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001003060820942245,
+      "loss": 0.2015,
+      "step": 43600
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00010008443950747599,
+      "loss": 0.2014,
+      "step": 43700
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.986257728051483e-05,
+      "loss": 0.2014,
+      "step": 43800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.964049783959082e-05,
+      "loss": 0.2012,
+      "step": 43900
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.94182036133291e-05,
+      "loss": 0.201,
+      "step": 44000
+    },
+    {
+      "epoch": 1.05,
+      "eval_runtime": 47.2136,
+      "eval_samples_per_second": 228.748,
+      "eval_steps_per_second": 7.159,
+      "step": 44000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.919569703270376e-05,
+      "loss": 0.1998,
+      "step": 44100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.89729805310111e-05,
+      "loss": 0.2004,
+      "step": 44200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.875005654384307e-05,
+      "loss": 0.2009,
+      "step": 44300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.852692750906071e-05,
+      "loss": 0.1999,
+      "step": 44400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.830359586676737e-05,
+      "loss": 0.1997,
+      "step": 44500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.808006405928215e-05,
+      "loss": 0.2006,
+      "step": 44600
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.785633453111306e-05,
+      "loss": 0.1999,
+      "step": 44700
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.763240972893037e-05,
+      "loss": 0.1992,
+      "step": 44800
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.740829210153984e-05,
+      "loss": 0.1991,
+      "step": 44900
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.718398409985593e-05,
+      "loss": 0.199,
+      "step": 45000
+    },
+    {
+      "epoch": 1.07,
+      "eval_runtime": 46.9221,
+      "eval_samples_per_second": 230.169,
+      "eval_steps_per_second": 7.203,
+      "step": 45000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.695948817687504e-05,
+      "loss": 0.1987,
+      "step": 45100
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.673480678764858e-05,
+      "loss": 0.1982,
+      "step": 45200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.650994238925626e-05,
+      "loss": 0.1989,
+      "step": 45300
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.628489744077911e-05,
+      "loss": 0.1985,
+      "step": 45400
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.60596744032726e-05,
+      "loss": 0.1981,
+      "step": 45500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.583427573973982e-05,
+      "loss": 0.1976,
+      "step": 45600
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.560870391510441e-05,
+      "loss": 0.1981,
+      "step": 45700
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.538296139618371e-05,
+      "loss": 0.1978,
+      "step": 45800
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.515705065166178e-05,
+      "loss": 0.1977,
+      "step": 45900
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.493097415206228e-05,
+      "loss": 0.1974,
+      "step": 46000
+    },
+    {
+      "epoch": 1.1,
+      "eval_runtime": 47.1161,
+      "eval_samples_per_second": 229.221,
+      "eval_steps_per_second": 7.174,
+      "step": 46000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.47047343697216e-05,
+      "loss": 0.1978,
+      "step": 46100
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.447833377876176e-05,
+      "loss": 0.1974,
+      "step": 46200
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.425177485506336e-05,
+      "loss": 0.1971,
+      "step": 46300
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.402506007623848e-05,
+      "loss": 0.1968,
+      "step": 46400
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.379819192160362e-05,
+      "loss": 0.1969,
+      "step": 46500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.357117287215258e-05,
+      "loss": 0.1966,
+      "step": 46600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.334400541052928e-05,
+      "loss": 0.1971,
+      "step": 46700
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.311669202100073e-05,
+      "loss": 0.1962,
+      "step": 46800
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.288923518942968e-05,
+      "loss": 0.1959,
+      "step": 46900
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.26616374032477e-05,
+      "loss": 0.1964,
+      "step": 47000
+    },
+    {
+      "epoch": 1.12,
+      "eval_runtime": 46.7963,
+      "eval_samples_per_second": 230.788,
+      "eval_steps_per_second": 7.223,
+      "step": 47000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.243390115142761e-05,
+      "loss": 0.196,
+      "step": 47100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.220602892445661e-05,
+      "loss": 0.1955,
+      "step": 47200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.197802321430889e-05,
+      "loss": 0.1958,
+      "step": 47300
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.174988651441833e-05,
+      "loss": 0.1951,
+      "step": 47400
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.152162131965137e-05,
+      "loss": 0.1954,
+      "step": 47500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.129323012627956e-05,
+      "loss": 0.1948,
+      "step": 47600
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.106471543195244e-05,
+      "loss": 0.1954,
+      "step": 47700
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.08360797356701e-05,
+      "loss": 0.1953,
+      "step": 47800
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.060732553775582e-05,
+      "loss": 0.1949,
+      "step": 47900
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.037845533982892e-05,
+      "loss": 0.1947,
+      "step": 48000
+    },
+    {
+      "epoch": 1.14,
+      "eval_runtime": 46.9646,
+      "eval_samples_per_second": 229.96,
+      "eval_steps_per_second": 7.197,
+      "step": 48000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.014947164477721e-05,
+      "loss": 0.1946,
+      "step": 48100
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.992037695672967e-05,
+      "loss": 0.1938,
+      "step": 48200
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.969117378102912e-05,
+      "loss": 0.1946,
+      "step": 48300
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 8.946186462420478e-05,
+      "loss": 0.1942,
+      "step": 48400
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.923245199394482e-05,
+      "loss": 0.1934,
+      "step": 48500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.900293839906903e-05,
+      "loss": 0.194,
+      "step": 48600
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.87733263495013e-05,
+      "loss": 0.1936,
+      "step": 48700
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.85436183562422e-05,
+      "loss": 0.1933,
+      "step": 48800
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 8.83138169313416e-05,
+      "loss": 0.1933,
+      "step": 48900
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.808392458787103e-05,
+      "loss": 0.1931,
+      "step": 49000
+    },
+    {
+      "epoch": 1.17,
+      "eval_runtime": 46.9712,
+      "eval_samples_per_second": 229.928,
+      "eval_steps_per_second": 7.196,
+      "step": 49000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.78539438398963e-05,
+      "loss": 0.1922,
+      "step": 49100
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.762387720245008e-05,
+      "loss": 0.1922,
+      "step": 49200
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 8.73937271915042e-05,
+      "loss": 0.1926,
+      "step": 49300
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.716349632394235e-05,
+      "loss": 0.1924,
+      "step": 49400
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.69331871175324e-05,
+      "loss": 0.1927,
+      "step": 49500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.67028020908989e-05,
+      "loss": 0.1924,
+      "step": 49600
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 8.647234376349565e-05,
+      "loss": 0.1921,
+      "step": 49700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.624181465557794e-05,
+      "loss": 0.1914,
+      "step": 49800
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.601121728817519e-05,
+      "loss": 0.1917,
+      "step": 49900
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.578055418306327e-05,
+      "loss": 0.1918,
+      "step": 50000
+    },
+    {
+      "epoch": 1.19,
+      "eval_runtime": 47.0452,
+      "eval_samples_per_second": 229.566,
+      "eval_steps_per_second": 7.185,
+      "step": 50000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 8.55498278627369e-05,
+      "loss": 0.1915,
+      "step": 50100
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.531904085038221e-05,
+      "loss": 0.1912,
+      "step": 50200
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.508819566984897e-05,
+      "loss": 0.1907,
+      "step": 50300
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.485729484562307e-05,
+      "loss": 0.1912,
+      "step": 50400
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 8.462634090279895e-05,
+      "loss": 0.1907,
+      "step": 50500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.439533636705194e-05,
+      "loss": 0.1912,
+      "step": 50600
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.416428376461061e-05,
+      "loss": 0.19,
+      "step": 50700
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.393318562222916e-05,
+      "loss": 0.1904,
+      "step": 50800
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.370204446715997e-05,
+      "loss": 0.1902,
+      "step": 50900
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.347086282712556e-05,
+      "loss": 0.191,
+      "step": 51000
+    },
+    {
+      "epoch": 1.21,
+      "eval_runtime": 46.8635,
+      "eval_samples_per_second": 230.456,
+      "eval_steps_per_second": 7.212,
+      "step": 51000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.323964323029136e-05,
+      "loss": 0.1896,
+      "step": 51100
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.300838820523784e-05,
+      "loss": 0.1903,
+      "step": 51200
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.277710028093289e-05,
+      "loss": 0.1895,
+      "step": 51300
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.254578198670421e-05,
+      "loss": 0.1897,
+      "step": 51400
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 8.231443585221157e-05,
+      "loss": 0.1895,
+      "step": 51500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 8.208306440741926e-05,
+      "loss": 0.1898,
+      "step": 51600
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 8.185167018256834e-05,
+      "loss": 0.1899,
+      "step": 51700
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 8.162025570814896e-05,
+      "loss": 0.1894,
+      "step": 51800
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 8.138882351487275e-05,
+      "loss": 0.1895,
+      "step": 51900
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 8.115737613364511e-05,
+      "loss": 0.1895,
+      "step": 52000
+    },
+    {
+      "epoch": 1.24,
+      "eval_runtime": 47.2771,
+      "eval_samples_per_second": 228.441,
+      "eval_steps_per_second": 7.149,
+      "step": 52000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 8.092591609553747e-05,
+      "loss": 0.1894,
+      "step": 52100
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 8.069444593175975e-05,
+      "loss": 0.1897,
+      "step": 52200
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 8.046296817363259e-05,
+      "loss": 0.1887,
+      "step": 52300
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 8.023148535255965e-05,
+      "loss": 0.1886,
+      "step": 52400
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 0.1885,
+      "step": 52500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 7.976851464744033e-05,
+      "loss": 0.1888,
+      "step": 52600
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.953703182636741e-05,
+      "loss": 0.1882,
+      "step": 52700
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.930555406824026e-05,
+      "loss": 0.1879,
+      "step": 52800
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.907408390446254e-05,
+      "loss": 0.1887,
+      "step": 52900
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.884262386635489e-05,
+      "loss": 0.1876,
+      "step": 53000
+    },
+    {
+      "epoch": 1.26,
+      "eval_runtime": 47.1816,
+      "eval_samples_per_second": 228.903,
+      "eval_steps_per_second": 7.164,
+      "step": 53000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 7.861117648512725e-05,
+      "loss": 0.1875,
+      "step": 53100
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.837974429185103e-05,
+      "loss": 0.1875,
+      "step": 53200
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.814832981743164e-05,
+      "loss": 0.1873,
+      "step": 53300
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.791693559258072e-05,
+      "loss": 0.1871,
+      "step": 53400
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 7.768556414778842e-05,
+      "loss": 0.1876,
+      "step": 53500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.74542180132958e-05,
+      "loss": 0.1871,
+      "step": 53600
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.72228997190671e-05,
+      "loss": 0.1877,
+      "step": 53700
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.699161179476217e-05,
+      "loss": 0.1864,
+      "step": 53800
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 7.676035676970863e-05,
+      "loss": 0.1864,
+      "step": 53900
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.652913717287443e-05,
+      "loss": 0.1862,
+      "step": 54000
+    },
+    {
+      "epoch": 1.29,
+      "eval_runtime": 46.9744,
+      "eval_samples_per_second": 229.912,
+      "eval_steps_per_second": 7.195,
+      "step": 54000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.629795553284005e-05,
+      "loss": 0.1868,
+      "step": 54100
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.606681437777081e-05,
+      "loss": 0.1867,
+      "step": 54200
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 7.583571623538939e-05,
+      "loss": 0.1858,
+      "step": 54300
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.560466363294806e-05,
+      "loss": 0.1865,
+      "step": 54400
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.537365909720104e-05,
+      "loss": 0.1859,
+      "step": 54500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.514270515437691e-05,
+      "loss": 0.1862,
+      "step": 54600
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 7.491180433015101e-05,
+      "loss": 0.1852,
+      "step": 54700
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.468095914961777e-05,
+      "loss": 0.1864,
+      "step": 54800
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.445017213726307e-05,
+      "loss": 0.1856,
+      "step": 54900
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 7.421944581693674e-05,
+      "loss": 0.1852,
+      "step": 55000
+    },
+    {
+      "epoch": 1.31,
+      "eval_runtime": 75.8452,
+      "eval_samples_per_second": 142.395,
+      "eval_steps_per_second": 4.456,
+      "step": 55000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.39887827118248e-05,
+      "loss": 0.1855,
+      "step": 55100
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.375818534442207e-05,
+      "loss": 0.1852,
+      "step": 55200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.352765623650435e-05,
+      "loss": 0.1858,
+      "step": 55300
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.329719790910108e-05,
+      "loss": 0.1842,
+      "step": 55400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.30668128824676e-05,
+      "loss": 0.185,
+      "step": 55500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 7.283650367605764e-05,
+      "loss": 0.1851,
+      "step": 55600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.260627280849581e-05,
+      "loss": 0.1842,
+      "step": 55700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.23761227975499e-05,
+      "loss": 0.1847,
+      "step": 55800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.21460561601037e-05,
+      "loss": 0.1849,
+      "step": 55900
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 7.191607541212897e-05,
+      "loss": 0.1848,
+      "step": 56000
+    },
+    {
+      "epoch": 1.02,
+      "eval_runtime": 47.6864,
+      "eval_samples_per_second": 226.48,
+      "eval_steps_per_second": 7.088,
+      "step": 56000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.168618306865838e-05,
+      "loss": 0.1848,
+      "step": 56100
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.145638164375779e-05,
+      "loss": 0.1842,
+      "step": 56200
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.122667365049869e-05,
+      "loss": 0.1846,
+      "step": 56300
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 7.099706160093098e-05,
+      "loss": 0.1834,
+      "step": 56400
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.076754800605516e-05,
+      "loss": 0.1837,
+      "step": 56500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.053813537579523e-05,
+      "loss": 0.1835,
+      "step": 56600
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.030882621897088e-05,
+      "loss": 0.1833,
+      "step": 56700
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 7.00796230432703e-05,
+      "loss": 0.1837,
+      "step": 56800
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.985052835522279e-05,
+      "loss": 0.1833,
+      "step": 56900
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.962154466017105e-05,
+      "loss": 0.1827,
+      "step": 57000
+    },
+    {
+      "epoch": 1.05,
+      "eval_runtime": 47.1326,
+      "eval_samples_per_second": 229.141,
+      "eval_steps_per_second": 7.171,
+      "step": 57000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.939267446224418e-05,
+      "loss": 0.1819,
+      "step": 57100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.91639202643299e-05,
+      "loss": 0.1834,
+      "step": 57200
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 6.893528456804756e-05,
+      "loss": 0.1836,
+      "step": 57300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.870676987372044e-05,
+      "loss": 0.1832,
+      "step": 57400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.847837868034861e-05,
+      "loss": 0.1833,
+      "step": 57500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.825011348558167e-05,
+      "loss": 0.1826,
+      "step": 57600
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 6.802197678569109e-05,
+      "loss": 0.1826,
+      "step": 57700
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.779397107554339e-05,
+      "loss": 0.1821,
+      "step": 57800
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.756609884857239e-05,
+      "loss": 0.1826,
+      "step": 57900
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.733836259675233e-05,
+      "loss": 0.1822,
+      "step": 58000
+    },
+    {
+      "epoch": 1.07,
+      "eval_runtime": 47.2136,
+      "eval_samples_per_second": 228.748,
+      "eval_steps_per_second": 7.159,
+      "step": 58000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 6.71107648105703e-05,
+      "loss": 0.1814,
+      "step": 58100
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.688330797899925e-05,
+      "loss": 0.1825,
+      "step": 58200
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.665599458947072e-05,
+      "loss": 0.182,
+      "step": 58300
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.642882712784742e-05,
+      "loss": 0.1821,
+      "step": 58400
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 6.620180807839639e-05,
+      "loss": 0.1819,
+      "step": 58500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.597493992376152e-05,
+      "loss": 0.1824,
+      "step": 58600
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.574822514493664e-05,
+      "loss": 0.1821,
+      "step": 58700
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.552166622123824e-05,
+      "loss": 0.1817,
+      "step": 58800
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 6.52952656302784e-05,
+      "loss": 0.181,
+      "step": 58900
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.506902584793773e-05,
+      "loss": 0.1814,
+      "step": 59000
+    },
+    {
+      "epoch": 1.1,
+      "eval_runtime": 47.2104,
+      "eval_samples_per_second": 228.763,
+      "eval_steps_per_second": 7.159,
+      "step": 59000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.484294934833822e-05,
+      "loss": 0.182,
+      "step": 59100
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.461703860381628e-05,
+      "loss": 0.1811,
+      "step": 59200
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.439129608489559e-05,
+      "loss": 0.1801,
+      "step": 59300
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 6.41657242602602e-05,
+      "loss": 0.1811,
+      "step": 59400
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.39403255967274e-05,
+      "loss": 0.1811,
+      "step": 59500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.371510255922088e-05,
+      "loss": 0.1811,
+      "step": 59600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.349005761074372e-05,
+      "loss": 0.1808,
+      "step": 59700
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 6.326519321235139e-05,
+      "loss": 0.1803,
+      "step": 59800
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.304051182312496e-05,
+      "loss": 0.1809,
+      "step": 59900
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.281601590014407e-05,
+      "loss": 0.1807,
+      "step": 60000
+    },
+    {
+      "epoch": 1.12,
+      "eval_runtime": 47.3036,
+      "eval_samples_per_second": 228.313,
+      "eval_steps_per_second": 7.145,
+      "step": 60000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.259170789846017e-05,
+      "loss": 0.1806,
+      "step": 60100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 6.236759027106965e-05,
+      "loss": 0.1803,
+      "step": 60200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.214366546888694e-05,
+      "loss": 0.1805,
+      "step": 60300
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.191993594071785e-05,
+      "loss": 0.1798,
+      "step": 60400
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.169640413323262e-05,
+      "loss": 0.1791,
+      "step": 60500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 6.147307249093929e-05,
+      "loss": 0.1793,
+      "step": 60600
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.124994345615693e-05,
+      "loss": 0.18,
+      "step": 60700
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.102701946898891e-05,
+      "loss": 0.1795,
+      "step": 60800
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.0804302967296225e-05,
+      "loss": 0.1791,
+      "step": 60900
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 6.058179638667089e-05,
+      "loss": 0.1798,
+      "step": 61000
+    },
+    {
+      "epoch": 1.14,
+      "eval_runtime": 47.2093,
+      "eval_samples_per_second": 228.768,
+      "eval_steps_per_second": 7.16,
+      "step": 61000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.035950216040917e-05,
+      "loss": 0.1793,
+      "step": 61100
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 6.0137422719485145e-05,
+      "loss": 0.1797,
+      "step": 61200
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.991556049252401e-05,
+      "loss": 0.1789,
+      "step": 61300
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.969391790577551e-05,
+      "loss": 0.1793,
+      "step": 61400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.947249738308747e-05,
+      "loss": 0.179,
+      "step": 61500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.925130134587924e-05,
+      "loss": 0.1785,
+      "step": 61600
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.903033221311528e-05,
+      "loss": 0.1787,
+      "step": 61700
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.880959240127858e-05,
+      "loss": 0.179,
+      "step": 61800
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.858908432434438e-05,
+      "loss": 0.1784,
+      "step": 61900
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.8368810393753684e-05,
+      "loss": 0.1789,
+      "step": 62000
+    },
+    {
+      "epoch": 1.17,
+      "eval_runtime": 47.3247,
+      "eval_samples_per_second": 228.211,
+      "eval_steps_per_second": 7.142,
+      "step": 62000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.814877301838688e-05,
+      "loss": 0.1783,
+      "step": 62100
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.7928974604537494e-05,
+      "loss": 0.1783,
+      "step": 62200
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.770941755588573e-05,
+      "loss": 0.1785,
+      "step": 62300
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.749010427347233e-05,
+      "loss": 0.1784,
+      "step": 62400
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.7271037155672156e-05,
+      "loss": 0.1777,
+      "step": 62500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.7052218598168154e-05,
+      "loss": 0.1786,
+      "step": 62600
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.6833650993925016e-05,
+      "loss": 0.1782,
+      "step": 62700
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.661533673316303e-05,
+      "loss": 0.1776,
+      "step": 62800
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.639727820333198e-05,
+      "loss": 0.178,
+      "step": 62900
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.617947778908498e-05,
+      "loss": 0.1782,
+      "step": 63000
+    },
+    {
+      "epoch": 1.19,
+      "eval_runtime": 47.2795,
+      "eval_samples_per_second": 228.429,
+      "eval_steps_per_second": 7.149,
+      "step": 63000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.596193787225254e-05,
+      "loss": 0.1771,
+      "step": 63100
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.574466083181624e-05,
+      "loss": 0.1777,
+      "step": 63200
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.552764904388305e-05,
+      "loss": 0.1773,
+      "step": 63300
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5310904881659116e-05,
+      "loss": 0.177,
+      "step": 63400
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5094430715423835e-05,
+      "loss": 0.1766,
+      "step": 63500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.487822891250406e-05,
+      "loss": 0.1771,
+      "step": 63600
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.4662301837247985e-05,
+      "loss": 0.177,
+      "step": 63700
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.4446651850999604e-05,
+      "loss": 0.1765,
+      "step": 63800
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.4231281312072544e-05,
+      "loss": 0.1774,
+      "step": 63900
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.401619257572453e-05,
+      "loss": 0.1766,
+      "step": 64000
+    },
+    {
+      "epoch": 1.21,
+      "eval_runtime": 47.4102,
+      "eval_samples_per_second": 227.799,
+      "eval_steps_per_second": 7.129,
+      "step": 64000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.3801387994131576e-05,
+      "loss": 0.1769,
+      "step": 64100
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.358686991636209e-05,
+      "loss": 0.1768,
+      "step": 64200
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.3372640688351476e-05,
+      "loss": 0.1767,
+      "step": 64300
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.315870265287618e-05,
+      "loss": 0.1762,
+      "step": 64400
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.294505814952835e-05,
+      "loss": 0.1771,
+      "step": 64500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.2731709514689995e-05,
+      "loss": 0.1759,
+      "step": 64600
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.25186590815076e-05,
+      "loss": 0.1759,
+      "step": 64700
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.2305909179866635e-05,
+      "loss": 0.1765,
+      "step": 64800
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.209346213636584e-05,
+      "loss": 0.1763,
+      "step": 64900
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.188132027429215e-05,
+      "loss": 0.1757,
+      "step": 65000
+    },
+    {
+      "epoch": 1.24,
+      "eval_runtime": 47.3383,
+      "eval_samples_per_second": 228.145,
+      "eval_steps_per_second": 7.14,
+      "step": 65000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.166948591359489e-05,
+      "loss": 0.1757,
+      "step": 65100
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.145796137086076e-05,
+      "loss": 0.176,
+      "step": 65200
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.124674895928823e-05,
+      "loss": 0.1759,
+      "step": 65300
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.103585098866237e-05,
+      "loss": 0.1758,
+      "step": 65400
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.082526976532968e-05,
+      "loss": 0.1754,
+      "step": 65500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.061500759217261e-05,
+      "loss": 0.1751,
+      "step": 65600
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.04050667685846e-05,
+      "loss": 0.1759,
+      "step": 65700
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.01954495904449e-05,
+      "loss": 0.1761,
+      "step": 65800
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.998615835009339e-05,
+      "loss": 0.1757,
+      "step": 65900
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.97771953363055e-05,
+      "loss": 0.1751,
+      "step": 66000
+    },
+    {
+      "epoch": 1.26,
+      "eval_runtime": 47.6107,
+      "eval_samples_per_second": 226.84,
+      "eval_steps_per_second": 7.099,
+      "step": 66000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.956856283426728e-05,
+      "loss": 0.1747,
+      "step": 66100
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.936026312555037e-05,
+      "loss": 0.1746,
+      "step": 66200
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.915229848808698e-05,
+      "loss": 0.1747,
+      "step": 66300
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.8944671196145136e-05,
+      "loss": 0.1744,
+      "step": 66400
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.8737383520303546e-05,
+      "loss": 0.1748,
+      "step": 66500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.853043772742709e-05,
+      "loss": 0.1748,
+      "step": 66600
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.832383608064172e-05,
+      "loss": 0.1746,
+      "step": 66700
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.811758083931005e-05,
+      "loss": 0.1754,
+      "step": 66800
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.791167425900632e-05,
+      "loss": 0.1744,
+      "step": 66900
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.770611859149185e-05,
+      "loss": 0.1742,
+      "step": 67000
+    },
+    {
+      "epoch": 1.29,
+      "eval_runtime": 47.4501,
+      "eval_samples_per_second": 227.608,
+      "eval_steps_per_second": 7.123,
+      "step": 67000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.7500916084690564e-05,
+      "loss": 0.174,
+      "step": 67100
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.729606898266411e-05,
+      "loss": 0.1742,
+      "step": 67200
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.709157952558768e-05,
+      "loss": 0.1743,
+      "step": 67300
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.688744994972514e-05,
+      "loss": 0.175,
+      "step": 67400
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.668368248740485e-05,
+      "loss": 0.1748,
+      "step": 67500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.6480279366995116e-05,
+      "loss": 0.1734,
+      "step": 67600
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.6277242812879914e-05,
+      "loss": 0.1736,
+      "step": 67700
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.607457504543447e-05,
+      "loss": 0.1739,
+      "step": 67800
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5872278281000955e-05,
+      "loss": 0.1739,
+      "step": 67900
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.567035473186444e-05,
+      "loss": 0.1743,
+      "step": 68000
+    },
+    {
+      "epoch": 1.31,
+      "eval_runtime": 47.6447,
+      "eval_samples_per_second": 226.678,
+      "eval_steps_per_second": 7.094,
+      "step": 68000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.546880660622845e-05,
+      "loss": 0.1737,
+      "step": 68100
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5267636108191036e-05,
+      "loss": 0.174,
+      "step": 68200
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.5066845437720555e-05,
+      "loss": 0.1735,
+      "step": 68300
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.4866436790631564e-05,
+      "loss": 0.1733,
+      "step": 68400
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.4666412358560955e-05,
+      "loss": 0.1733,
+      "step": 68500
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.4466774328943796e-05,
+      "loss": 0.1729,
+      "step": 68600
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.426752488498972e-05,
+      "loss": 0.1735,
+      "step": 68700
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.406866620565862e-05,
+      "loss": 0.173,
+      "step": 68800
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.3870200465637164e-05,
+      "loss": 0.1732,
+      "step": 68900
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.3672129835314955e-05,
+      "loss": 0.1727,
+      "step": 69000
+    },
+    {
+      "epoch": 1.33,
+      "eval_runtime": 47.5417,
+      "eval_samples_per_second": 227.169,
+      "eval_steps_per_second": 7.11,
+      "step": 69000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.347445648076057e-05,
+      "loss": 0.1738,
+      "step": 69100
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.327718256369826e-05,
+      "loss": 0.1725,
+      "step": 69200
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.3080310241483885e-05,
+      "loss": 0.1731,
+      "step": 69300
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.2883841667081675e-05,
+      "loss": 0.1731,
+      "step": 69400
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.268777898904044e-05,
+      "loss": 0.1726,
+      "step": 69500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.2492124351470214e-05,
+      "loss": 0.1723,
+      "step": 69600
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.2296879894018835e-05,
+      "loss": 0.1727,
+      "step": 69700
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.210204775184834e-05,
+      "loss": 0.1723,
+      "step": 69800
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.190763005561186e-05,
+      "loss": 0.172,
+      "step": 69900
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.171362893143013e-05,
+      "loss": 0.1724,
+      "step": 70000
+    },
+    {
+      "epoch": 1.36,
+      "eval_runtime": 47.6533,
+      "eval_samples_per_second": 226.637,
+      "eval_steps_per_second": 7.093,
+      "step": 70000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.1520046500868384e-05,
+      "loss": 0.1724,
+      "step": 70100
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.1326884880913074e-05,
+      "loss": 0.1721,
+      "step": 70200
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.1134146183948724e-05,
+      "loss": 0.1723,
+      "step": 70300
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.0941832517734885e-05,
+      "loss": 0.1717,
+      "step": 70400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.0749945985382915e-05,
+      "loss": 0.1717,
+      "step": 70500
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.0558488685333235e-05,
+      "loss": 0.1713,
+      "step": 70600
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.036746271133223e-05,
+      "loss": 0.1724,
+      "step": 70700
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.0176870152409324e-05,
+      "loss": 0.1708,
+      "step": 70800
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.998671309285417e-05,
+      "loss": 0.1717,
+      "step": 70900
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.979699361219395e-05,
+      "loss": 0.1706,
+      "step": 71000
+    },
+    {
+      "epoch": 1.38,
+      "eval_runtime": 47.7292,
+      "eval_samples_per_second": 226.277,
+      "eval_steps_per_second": 7.082,
+      "step": 71000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.960771378517049e-05,
+      "loss": 0.171,
+      "step": 71100
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.941887568171766e-05,
+      "loss": 0.1708,
+      "step": 71200
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.923048136693873e-05,
+      "loss": 0.1717,
+      "step": 71300
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.904253290108369e-05,
+      "loss": 0.1716,
+      "step": 71400
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.885503233952689e-05,
+      "loss": 0.1706,
+      "step": 71500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.86679817327444e-05,
+      "loss": 0.171,
+      "step": 71600
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.848138312629171e-05,
+      "loss": 0.1702,
+      "step": 71700
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.8295238560781317e-05,
+      "loss": 0.1706,
+      "step": 71800
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.810955007186029e-05,
+      "loss": 0.1708,
+      "step": 71900
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.792431969018824e-05,
+      "loss": 0.1709,
+      "step": 72000
+    },
+    {
+      "epoch": 1.4,
+      "eval_runtime": 47.6465,
+      "eval_samples_per_second": 226.669,
+      "eval_steps_per_second": 7.094,
+      "step": 72000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.7739549441414945e-05,
+      "loss": 0.1703,
+      "step": 72100
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.755524134615825e-05,
+      "loss": 0.171,
+      "step": 72200
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.7371397419981925e-05,
+      "loss": 0.1706,
+      "step": 72300
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.7188019673373706e-05,
+      "loss": 0.1707,
+      "step": 72400
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.700511011172325e-05,
+      "loss": 0.1706,
+      "step": 72500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.682267073530023e-05,
+      "loss": 0.1703,
+      "step": 72600
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.664070353923245e-05,
+      "loss": 0.1698,
+      "step": 72700
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.645921051348396e-05,
+      "loss": 0.1705,
+      "step": 72800
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.627819364283345e-05,
+      "loss": 0.1708,
+      "step": 72900
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.6097654906852405e-05,
+      "loss": 0.1706,
+      "step": 73000
+    },
+    {
+      "epoch": 1.43,
+      "eval_runtime": 47.2921,
+      "eval_samples_per_second": 228.368,
+      "eval_steps_per_second": 7.147,
+      "step": 73000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.591759627988353e-05,
+      "loss": 0.17,
+      "step": 73100
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.573801973101913e-05,
+      "loss": 0.1702,
+      "step": 73200
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.5558927224079534e-05,
+      "loss": 0.1702,
+      "step": 73300
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.5380320717591716e-05,
+      "loss": 0.17,
+      "step": 73400
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.5202202164767836e-05,
+      "loss": 0.17,
+      "step": 73500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.5024573513483864e-05,
+      "loss": 0.1706,
+      "step": 73600
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.484743670625822e-05,
+      "loss": 0.1701,
+      "step": 73700
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.467079368023068e-05,
+      "loss": 0.1691,
+      "step": 73800
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.449464636714107e-05,
+      "loss": 0.1698,
+      "step": 73900
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.431899669330819e-05,
+      "loss": 0.1703,
+      "step": 74000
+    },
+    {
+      "epoch": 1.45,
+      "eval_runtime": 47.7148,
+      "eval_samples_per_second": 226.345,
+      "eval_steps_per_second": 7.084,
+      "step": 74000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.4143846579608744e-05,
+      "loss": 0.1688,
+      "step": 74100
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.396919794145629e-05,
+      "loss": 0.169,
+      "step": 74200
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.3795052688780345e-05,
+      "loss": 0.1691,
+      "step": 74300
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.362141272600552e-05,
+      "loss": 0.1695,
+      "step": 74400
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.3448279952030615e-05,
+      "loss": 0.1692,
+      "step": 74500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.327565626020793e-05,
+      "loss": 0.1697,
+      "step": 74600
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.3103543538322455e-05,
+      "loss": 0.1694,
+      "step": 74700
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.293194366857137e-05,
+      "loss": 0.1686,
+      "step": 74800
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.276085852754336e-05,
+      "loss": 0.1686,
+      "step": 74900
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.259028998619814e-05,
+      "loss": 0.1688,
+      "step": 75000
+    },
+    {
+      "epoch": 1.48,
+      "eval_runtime": 47.4078,
+      "eval_samples_per_second": 227.811,
+      "eval_steps_per_second": 7.13,
+      "step": 75000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.2420239909845894e-05,
+      "loss": 0.1688,
+      "step": 75100
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.2250710158127045e-05,
+      "loss": 0.1692,
+      "step": 75200
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.2081702584991786e-05,
+      "loss": 0.1692,
+      "step": 75300
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.191321903867988e-05,
+      "loss": 0.1689,
+      "step": 75400
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.174526136170039e-05,
+      "loss": 0.1691,
+      "step": 75500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.157783139081155e-05,
+      "loss": 0.1686,
+      "step": 75600
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.141093095700072e-05,
+      "loss": 0.1687,
+      "step": 75700
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.1244561885464244e-05,
+      "loss": 0.1683,
+      "step": 75800
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.107872599558769e-05,
+      "loss": 0.1687,
+      "step": 75900
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.0913425100925795e-05,
+      "loss": 0.1685,
+      "step": 76000
+    },
+    {
+      "epoch": 1.5,
+      "eval_runtime": 47.5367,
+      "eval_samples_per_second": 227.193,
+      "eval_steps_per_second": 7.11,
+      "step": 76000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.0748661009182616e-05,
+      "loss": 0.1684,
+      "step": 76100
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.0584435522191896e-05,
+      "loss": 0.1684,
+      "step": 76200
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.0420750435897183e-05,
+      "loss": 0.1684,
+      "step": 76300
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.025760754033246e-05,
+      "loss": 0.1679,
+      "step": 76400
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.0095008619602206e-05,
+      "loss": 0.1676,
+      "step": 76500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.993295545186223e-05,
+      "loss": 0.1685,
+      "step": 76600
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.977144980929996e-05,
+      "loss": 0.1681,
+      "step": 76700
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.961049345811523e-05,
+      "loss": 0.1685,
+      "step": 76800
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.945008815850097e-05,
+      "loss": 0.1679,
+      "step": 76900
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.929023566462377e-05,
+      "loss": 0.1682,
+      "step": 77000
+    },
+    {
+      "epoch": 1.52,
+      "eval_runtime": 47.5934,
+      "eval_samples_per_second": 226.922,
+      "eval_steps_per_second": 7.102,
+      "step": 77000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.9130937724604947e-05,
+      "loss": 0.1678,
+      "step": 77100
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.8972196080501208e-05,
+      "loss": 0.1678,
+      "step": 77200
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.8814012468285748e-05,
+      "loss": 0.1682,
+      "step": 77300
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.865638861782922e-05,
+      "loss": 0.1678,
+      "step": 77400
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.849932625288079e-05,
+      "loss": 0.1681,
+      "step": 77500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.8342827091049336e-05,
+      "loss": 0.1678,
+      "step": 77600
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.8186892843784587e-05,
+      "loss": 0.1677,
+      "step": 77700
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.803152521635851e-05,
+      "loss": 0.1679,
+      "step": 77800
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7876725907846578e-05,
+      "loss": 0.1676,
+      "step": 77900
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7722496611109243e-05,
+      "loss": 0.167,
+      "step": 78000
+    },
+    {
+      "epoch": 1.55,
+      "eval_runtime": 47.8007,
+      "eval_samples_per_second": 225.938,
+      "eval_steps_per_second": 7.071,
+      "step": 78000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7568839012773365e-05,
+      "loss": 0.1673,
+      "step": 78100
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7415754793213826e-05,
+      "loss": 0.1676,
+      "step": 78200
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.7263245626535116e-05,
+      "loss": 0.1673,
+      "step": 78300
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.7111313180553077e-05,
+      "loss": 0.1673,
+      "step": 78400
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.6959959116776587e-05,
+      "loss": 0.1663,
+      "step": 78500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.6809185090389406e-05,
+      "loss": 0.1674,
+      "step": 78600
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.6658992750232167e-05,
+      "loss": 0.1666,
+      "step": 78700
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6509383738784218e-05,
+      "loss": 0.1671,
+      "step": 78800
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6360359692145757e-05,
+      "loss": 0.1669,
+      "step": 78900
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6211922240019883e-05,
+      "loss": 0.1671,
+      "step": 79000
+    },
+    {
+      "epoch": 1.57,
+      "eval_runtime": 47.5708,
+      "eval_samples_per_second": 227.03,
+      "eval_steps_per_second": 7.105,
+      "step": 79000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.6064073005694758e-05,
+      "loss": 0.1669,
+      "step": 79100
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.591681360602595e-05,
+      "loss": 0.1665,
+      "step": 79200
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.577014565141866e-05,
+      "loss": 0.1666,
+      "step": 79300
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.562407074581014e-05,
+      "loss": 0.1663,
+      "step": 79400
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.5478590486652137e-05,
+      "loss": 0.1667,
+      "step": 79500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.533370646489347e-05,
+      "loss": 0.1665,
+      "step": 79600
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.5189420264962586e-05,
+      "loss": 0.1658,
+      "step": 79700
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.504573346475026e-05,
+      "loss": 0.1665,
+      "step": 79800
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.4902647635592324e-05,
+      "loss": 0.1666,
+      "step": 79900
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.476016434225246e-05,
+      "loss": 0.1662,
+      "step": 80000
+    },
+    {
+      "epoch": 1.6,
+      "eval_runtime": 47.6449,
+      "eval_samples_per_second": 226.677,
+      "eval_steps_per_second": 7.094,
+      "step": 80000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.461828514290513e-05,
+      "loss": 0.1663,
+      "step": 80100
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.447701158911855e-05,
+      "loss": 0.1664,
+      "step": 80200
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4336345225837658e-05,
+      "loss": 0.1664,
+      "step": 80300
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.4196287591367296e-05,
+      "loss": 0.1656,
+      "step": 80400
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.405684021735527e-05,
+      "loss": 0.166,
+      "step": 80500
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3918004628775736e-05,
+      "loss": 0.166,
+      "step": 80600
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.3779782343912463e-05,
+      "loss": 0.1656,
+      "step": 80700
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.364217487434221e-05,
+      "loss": 0.1665,
+      "step": 80800
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3505183724918196e-05,
+      "loss": 0.1653,
+      "step": 80900
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.3368810393753687e-05,
+      "loss": 0.166,
+      "step": 81000
+    },
+    {
+      "epoch": 1.62,
+      "eval_runtime": 47.4693,
+      "eval_samples_per_second": 227.516,
+      "eval_steps_per_second": 7.12,
+      "step": 81000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.32330563722056e-05,
+      "loss": 0.1653,
+      "step": 81100
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.309792314485815e-05,
+      "loss": 0.1659,
+      "step": 81200
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2963412189506695e-05,
+      "loss": 0.1652,
+      "step": 81300
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.282952497714145e-05,
+      "loss": 0.1658,
+      "step": 81400
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2696262971931538e-05,
+      "loss": 0.1649,
+      "step": 81500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.2563627631208887e-05,
+      "loss": 0.1657,
+      "step": 81600
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2431620405452336e-05,
+      "loss": 0.1656,
+      "step": 81700
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.230024273827179e-05,
+      "loss": 0.1655,
+      "step": 81800
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.216949606639231e-05,
+      "loss": 0.1659,
+      "step": 81900
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.2039381819638596e-05,
+      "loss": 0.1649,
+      "step": 82000
+    },
+    {
+      "epoch": 1.64,
+      "eval_runtime": 47.7941,
+      "eval_samples_per_second": 225.969,
+      "eval_steps_per_second": 7.072,
+      "step": 82000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1909901420919184e-05,
+      "loss": 0.1655,
+      "step": 82100
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1781056286210997e-05,
+      "loss": 0.1653,
+      "step": 82200
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1652847824543744e-05,
+      "loss": 0.1651,
+      "step": 82300
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1525277437984636e-05,
+      "loss": 0.165,
+      "step": 82400
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.1398346521623e-05,
+      "loss": 0.1652,
+      "step": 82500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1272056463554978e-05,
+      "loss": 0.1645,
+      "step": 82600
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.114640864486845e-05,
+      "loss": 0.165,
+      "step": 82700
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.1021404439627775e-05,
+      "loss": 0.165,
+      "step": 82800
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.089704521485896e-05,
+      "loss": 0.1652,
+      "step": 82900
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.0773332330534513e-05,
+      "loss": 0.1655,
+      "step": 83000
+    },
+    {
+      "epoch": 1.67,
+      "eval_runtime": 47.9606,
+      "eval_samples_per_second": 225.185,
+      "eval_steps_per_second": 7.047,
+      "step": 83000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.0650267139558772e-05,
+      "loss": 0.1651,
+      "step": 83100
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.052785098775293e-05,
+      "loss": 0.1637,
+      "step": 83200
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.04060852138404e-05,
+      "loss": 0.1651,
+      "step": 83300
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.028497114943219e-05,
+      "loss": 0.1646,
+      "step": 83400
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.0164510119012263e-05,
+      "loss": 0.1652,
+      "step": 83500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.0044703439923217e-05,
+      "loss": 0.1649,
+      "step": 83600
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.9925552422351654e-05,
+      "loss": 0.1652,
+      "step": 83700
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.9807058369314016e-05,
+      "loss": 0.1644,
+      "step": 83800
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.968922257664231e-05,
+      "loss": 0.1647,
+      "step": 83900
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.9572046332969825e-05,
+      "loss": 0.1638,
+      "step": 84000
+    },
+    {
+      "epoch": 1.69,
+      "eval_runtime": 47.6904,
+      "eval_samples_per_second": 226.46,
+      "eval_steps_per_second": 7.087,
+      "step": 84000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.945553091971727e-05,
+      "loss": 0.1646,
+      "step": 84100
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.933967761107847e-05,
+      "loss": 0.165,
+      "step": 84200
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.9224487674006694e-05,
+      "loss": 0.164,
+      "step": 84300
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.9109962368200602e-05,
+      "loss": 0.1646,
+      "step": 84400
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.8996102946090586e-05,
+      "loss": 0.1647,
+      "step": 84500
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.888291065282509e-05,
+      "loss": 0.1642,
+      "step": 84600
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8770386726256865e-05,
+      "loss": 0.1634,
+      "step": 84700
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8658532396929565e-05,
+      "loss": 0.1638,
+      "step": 84800
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8547348888064178e-05,
+      "loss": 0.1642,
+      "step": 84900
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.8436837415545772e-05,
+      "loss": 0.1646,
+      "step": 85000
+    },
+    {
+      "epoch": 1.71,
+      "eval_runtime": 47.7791,
+      "eval_samples_per_second": 226.04,
+      "eval_steps_per_second": 7.074,
+      "step": 85000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8326999187910095e-05,
+      "loss": 0.1625,
+      "step": 85100
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.8217835406330415e-05,
+      "loss": 0.1622,
+      "step": 85200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.810934726460436e-05,
+      "loss": 0.1628,
+      "step": 85300
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.800153594914084e-05,
+      "loss": 0.1613,
+      "step": 85400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.7894402638947176e-05,
+      "loss": 0.1622,
+      "step": 85500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.778794850561604e-05,
+      "loss": 0.1622,
+      "step": 85600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7682174713312805e-05,
+      "loss": 0.1615,
+      "step": 85700
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.75770824187627e-05,
+      "loss": 0.1621,
+      "step": 85800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.747267277123821e-05,
+      "loss": 0.1623,
+      "step": 85900
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.7368946912546556e-05,
+      "loss": 0.1622,
+      "step": 86000
+    },
+    {
+      "epoch": 2.02,
+      "eval_runtime": 45.2924,
+      "eval_samples_per_second": 238.45,
+      "eval_steps_per_second": 7.463,
+      "step": 86000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.726590597701708e-05,
+      "loss": 0.1623,
+      "step": 86100
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.7163551091488952e-05,
+      "loss": 0.1619,
+      "step": 86200
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.7061883375298788e-05,
+      "loss": 0.1622,
+      "step": 86300
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.6960903940268456e-05,
+      "loss": 0.1613,
+      "step": 86400
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6860613890692876e-05,
+      "loss": 0.1615,
+      "step": 86500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6761014323327962e-05,
+      "loss": 0.1613,
+      "step": 86600
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6662106327378645e-05,
+      "loss": 0.1612,
+      "step": 86700
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.6563890984486884e-05,
+      "loss": 0.1617,
+      "step": 86800
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6466369368719955e-05,
+      "loss": 0.1614,
+      "step": 86900
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6369542546558626e-05,
+      "loss": 0.1608,
+      "step": 87000
+    },
+    {
+      "epoch": 2.05,
+      "eval_runtime": 45.3963,
+      "eval_samples_per_second": 237.905,
+      "eval_steps_per_second": 7.446,
+      "step": 87000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.6273411576885517e-05,
+      "loss": 0.1601,
+      "step": 87100
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.617797751097349e-05,
+      "loss": 0.1617,
+      "step": 87200
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.608324139247421e-05,
+      "loss": 0.1618,
+      "step": 87300
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5989204257406693e-05,
+      "loss": 0.1616,
+      "step": 87400
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5895867134145974e-05,
+      "loss": 0.1618,
+      "step": 87500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5803231043411912e-05,
+      "loss": 0.1611,
+      "step": 87600
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.5711296998257902e-05,
+      "loss": 0.1611,
+      "step": 87700
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.562006600405996e-05,
+      "loss": 0.1608,
+      "step": 87800
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.5529539058505624e-05,
+      "loss": 0.1612,
+      "step": 87900
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.543971715158307e-05,
+      "loss": 0.1611,
+      "step": 88000
+    },
+    {
+      "epoch": 2.07,
+      "eval_runtime": 45.4783,
+      "eval_samples_per_second": 237.476,
+      "eval_steps_per_second": 7.432,
+      "step": 88000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.535060126557028e-05,
+      "loss": 0.1601,
+      "step": 88100
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5262192375024284e-05,
+      "loss": 0.1615,
+      "step": 88200
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.5174491446770566e-05,
+      "loss": 0.161,
+      "step": 88300
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.508749943989242e-05,
+      "loss": 0.1612,
+      "step": 88400
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.500121730572051e-05,
+      "loss": 0.161,
+      "step": 88500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.4915645987822406e-05,
+      "loss": 0.1613,
+      "step": 88600
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.4830786421992347e-05,
+      "loss": 0.1611,
+      "step": 88700
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.4746639536240942e-05,
+      "loss": 0.161,
+      "step": 88800
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.4663206250785055e-05,
+      "loss": 0.1605,
+      "step": 88900
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.4580487478037748e-05,
+      "loss": 0.1609,
+      "step": 89000
+    },
+    {
+      "epoch": 2.1,
+      "eval_runtime": 45.5774,
+      "eval_samples_per_second": 236.96,
+      "eval_steps_per_second": 7.416,
+      "step": 89000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.4498484122598232e-05,
+      "loss": 0.1615,
+      "step": 89100
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.4417197081242083e-05,
+      "loss": 0.1605,
+      "step": 89200
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.433662724291136e-05,
+      "loss": 0.1596,
+      "step": 89300
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.4256775488704904e-05,
+      "loss": 0.1608,
+      "step": 89400
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4177642691868717e-05,
+      "loss": 0.1608,
+      "step": 89500
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4099229717786368e-05,
+      "loss": 0.1609,
+      "step": 89600
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.4021537423969588e-05,
+      "loss": 0.1607,
+      "step": 89700
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.3944566660048863e-05,
+      "loss": 0.1603,
+      "step": 89800
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.3868318267764128e-05,
+      "loss": 0.1608,
+      "step": 89900
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.3792793080955574e-05,
+      "loss": 0.1607,
+      "step": 90000
+    },
+    {
+      "epoch": 2.12,
+      "eval_runtime": 45.4935,
+      "eval_samples_per_second": 237.397,
+      "eval_steps_per_second": 7.43,
+      "step": 90000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.3717991925554562e-05,
+      "loss": 0.1608,
+      "step": 90100
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.3643915619574529e-05,
+      "loss": 0.1604,
+      "step": 90200
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.35705649731021e-05,
+      "loss": 0.1607,
+      "step": 90300
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.3497940788288195e-05,
+      "loss": 0.1602,
+      "step": 90400
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.3426043859339253e-05,
+      "loss": 0.1595,
+      "step": 90500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.3354874972508582e-05,
+      "loss": 0.1598,
+      "step": 90600
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.3284434906087695e-05,
+      "loss": 0.1602,
+      "step": 90700
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.3214724430397915e-05,
+      "loss": 0.1599,
+      "step": 90800
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.314574430778182e-05,
+      "loss": 0.1598,
+      "step": 90900
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.3077495292594966e-05,
+      "loss": 0.1604,
+      "step": 91000
+    },
+    {
+      "epoch": 2.14,
+      "eval_runtime": 45.6099,
+      "eval_samples_per_second": 236.791,
+      "eval_steps_per_second": 7.411,
+      "step": 91000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.3009978131197669e-05,
+      "loss": 0.16,
+      "step": 91100
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2943193561946762e-05,
+      "loss": 0.1604,
+      "step": 91200
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2877142315187628e-05,
+      "loss": 0.1597,
+      "step": 91300
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.28118251132461e-05,
+      "loss": 0.1601,
+      "step": 91400
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.274724267042063e-05,
+      "loss": 0.1599,
+      "step": 91500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2683395692974472e-05,
+      "loss": 0.1595,
+      "step": 91600
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2620284879127947e-05,
+      "loss": 0.1598,
+      "step": 91700
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2557910919050803e-05,
+      "loss": 0.1602,
+      "step": 91800
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2496274494854666e-05,
+      "loss": 0.1596,
+      "step": 91900
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.24353762805856e-05,
+      "loss": 0.1601,
+      "step": 92000
+    },
+    {
+      "epoch": 2.17,
+      "eval_runtime": 45.5979,
+      "eval_samples_per_second": 236.853,
+      "eval_steps_per_second": 7.413,
+      "step": 92000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2375216942216713e-05,
+      "loss": 0.1596,
+      "step": 92100
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2315797137640906e-05,
+      "loss": 0.1598,
+      "step": 92200
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.225711751666363e-05,
+      "loss": 0.16,
+      "step": 92300
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2199178720995825e-05,
+      "loss": 0.1598,
+      "step": 92400
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2141981384246874e-05,
+      "loss": 0.1593,
+      "step": 92500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2085526131917685e-05,
+      "loss": 0.1602,
+      "step": 92600
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2029813581393866e-05,
+      "loss": 0.1598,
+      "step": 92700
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.197484434193893e-05,
+      "loss": 0.1593,
+      "step": 92800
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.192061901468768e-05,
+      "loss": 0.1597,
+      "step": 92900
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.1867138192639601e-05,
+      "loss": 0.16,
+      "step": 93000
+    },
+    {
+      "epoch": 2.19,
+      "eval_runtime": 45.5858,
+      "eval_samples_per_second": 236.916,
+      "eval_steps_per_second": 7.415,
+      "step": 93000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.1814402460652382e-05,
+      "loss": 0.159,
+      "step": 93100
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.176241239543558e-05,
+      "loss": 0.1597,
+      "step": 93200
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.171116856554418e-05,
+      "loss": 0.1594,
+      "step": 93300
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.1660671531372517e-05,
+      "loss": 0.1591,
+      "step": 93400
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.1610921845148052e-05,
+      "loss": 0.1587,
+      "step": 93500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.156192005092539e-05,
+      "loss": 0.1593,
+      "step": 93600
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1513666684580308e-05,
+      "loss": 0.1593,
+      "step": 93700
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1466162273803876e-05,
+      "loss": 0.1587,
+      "step": 93800
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1419407338096732e-05,
+      "loss": 0.1598,
+      "step": 93900
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.1373402388763346e-05,
+      "loss": 0.159,
+      "step": 94000
+    },
+    {
+      "epoch": 2.21,
+      "eval_runtime": 45.6844,
+      "eval_samples_per_second": 236.404,
+      "eval_steps_per_second": 7.399,
+      "step": 94000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.1328147928906494e-05,
+      "loss": 0.1594,
+      "step": 94100
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.1283644453421678e-05,
+      "loss": 0.1593,
+      "step": 94200
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.1239892448991798e-05,
+      "loss": 0.1593,
+      "step": 94300
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.1196892394081743e-05,
+      "loss": 0.1588,
+      "step": 94400
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1154644758933235e-05,
+      "loss": 0.1598,
+      "step": 94500
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1113150005559644e-05,
+      "loss": 0.1587,
+      "step": 94600
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1072408587740942e-05,
+      "loss": 0.1589,
+      "step": 94700
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.1032420951018755e-05,
+      "loss": 0.1594,
+      "step": 94800
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.0993187532691458e-05,
+      "loss": 0.1593,
+      "step": 94900
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.0954708761809438e-05,
+      "loss": 0.1588,
+      "step": 95000
+    },
+    {
+      "epoch": 2.24,
+      "eval_runtime": 45.7912,
+      "eval_samples_per_second": 235.853,
+      "eval_steps_per_second": 7.381,
+      "step": 95000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.091698505917036e-05,
+      "loss": 0.1587,
+      "step": 95100
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.0880016837314599e-05,
+      "loss": 0.1592,
+      "step": 95200
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.084380450052071e-05,
+      "loss": 0.159,
+      "step": 95300
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.0808348444801e-05,
+      "loss": 0.1592,
+      "step": 95400
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.0773649057897206e-05,
+      "loss": 0.1588,
+      "step": 95500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.073970671927628e-05,
+      "loss": 0.1585,
+      "step": 95600
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.0706521800126198e-05,
+      "loss": 0.1593,
+      "step": 95700
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0674094663351906e-05,
+      "loss": 0.1595,
+      "step": 95800
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0642425663571383e-05,
+      "loss": 0.1593,
+      "step": 95900
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0611515147111736e-05,
+      "loss": 0.1587,
+      "step": 96000
+    },
+    {
+      "epoch": 2.26,
+      "eval_runtime": 45.5789,
+      "eval_samples_per_second": 236.952,
+      "eval_steps_per_second": 7.416,
+      "step": 96000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.0581363452005424e-05,
+      "loss": 0.1583,
+      "step": 96100
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.0551970907986557e-05,
+      "loss": 0.1584,
+      "step": 96200
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.0523337836487271e-05,
+      "loss": 0.1585,
+      "step": 96300
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.0495464550634267e-05,
+      "loss": 0.1583,
+      "step": 96400
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.046835135524533e-05,
+      "loss": 0.1587,
+      "step": 96500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.044199854682601e-05,
+      "loss": 0.1588,
+      "step": 96600
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.0416406413566414e-05,
+      "loss": 0.1586,
+      "step": 96700
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.0391575235337991e-05,
+      "loss": 0.1596,
+      "step": 96800
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.0367505283690547e-05,
+      "loss": 0.1585,
+      "step": 96900
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.0344196821849202e-05,
+      "loss": 0.1584,
+      "step": 97000
+    },
+    {
+      "epoch": 2.29,
+      "eval_runtime": 45.8042,
+      "eval_samples_per_second": 235.786,
+      "eval_steps_per_second": 7.379,
+      "step": 97000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.032165010471157e-05,
+      "loss": 0.1582,
+      "step": 97100
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.0299865378844936e-05,
+      "loss": 0.1586,
+      "step": 97200
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.0278842882483569e-05,
+      "loss": 0.1587,
+      "step": 97300
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.025858284552612e-05,
+      "loss": 0.1594,
+      "step": 97400
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.023908548953311e-05,
+      "loss": 0.1593,
+      "step": 97500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.02203510277245e-05,
+      "loss": 0.158,
+      "step": 97600
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.0202379664977364e-05,
+      "loss": 0.1582,
+      "step": 97700
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.018517159782365e-05,
+      "loss": 0.1586,
+      "step": 97800
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0168727014448004e-05,
+      "loss": 0.1586,
+      "step": 97900
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0153046094685783e-05,
+      "loss": 0.1591,
+      "step": 98000
+    },
+    {
+      "epoch": 2.31,
+      "eval_runtime": 45.7753,
+      "eval_samples_per_second": 235.935,
+      "eval_steps_per_second": 7.384,
+      "step": 98000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0138129010020992e-05,
+      "loss": 0.1586,
+      "step": 98100
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.0123975923584488e-05,
+      "loss": 0.1588,
+      "step": 98200
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.0110586990152152e-05,
+      "loss": 0.1585,
+      "step": 98300
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.0097962356143219e-05,
+      "loss": 0.1583,
+      "step": 98400
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.0086102159618668e-05,
+      "loss": 0.1584,
+      "step": 98500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.0075006530279694e-05,
+      "loss": 0.1579,
+      "step": 98600
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0064675589466339e-05,
+      "loss": 0.1586,
+      "step": 98700
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0055109450156098e-05,
+      "loss": 0.1583,
+      "step": 98800
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0046308216962759e-05,
+      "loss": 0.1585,
+      "step": 98900
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.0038271986135177e-05,
+      "loss": 0.1581,
+      "step": 99000
+    },
+    {
+      "epoch": 2.33,
+      "eval_runtime": 45.6993,
+      "eval_samples_per_second": 236.327,
+      "eval_steps_per_second": 7.396,
+      "step": 99000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0031000845556304e-05,
+      "loss": 0.1592,
+      "step": 99100
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0024494874742152e-05,
+      "loss": 0.158,
+      "step": 99200
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0018754144840986e-05,
+      "loss": 0.1586,
+      "step": 99300
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.0013778718632507e-05,
+      "loss": 0.1585,
+      "step": 99400
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.000956865052717e-05,
+      "loss": 0.1582,
+      "step": 99500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0006123986565623e-05,
+      "loss": 0.1579,
+      "step": 99600
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.0003444764418138e-05,
+      "loss": 0.1585,
+      "step": 99700
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.000153101338428e-05,
+      "loss": 0.1581,
+      "step": 99800
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.00003827543925e-05,
+      "loss": 0.1578,
+      "step": 99900
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1e-05,
+      "loss": 0.1583,
+      "step": 100000
+    },
+    {
+      "epoch": 2.36,
+      "eval_runtime": 45.8451,
+      "eval_samples_per_second": 235.576,
+      "eval_steps_per_second": 7.373,
+      "step": 100000
+    }
+  ],
+  "max_steps": 100000,
+  "num_train_epochs": 3,
+  "total_flos": 7.009987307584106e+21,
+  "trial_name": null,
+  "trial_params": null
+}