diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6205 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.389467,
+  "global_step": 1000000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9820000000000003e-06,
+      "loss": 82.5404,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98e-06,
+      "loss": 57.4563,
+      "step": 2000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.98e-06,
+      "loss": 41.077,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.978e-06,
+      "loss": 30.4085,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.976e-06,
+      "loss": 23.8433,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1976000000000002e-05,
+      "loss": 18.0328,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3976e-05,
+      "loss": 14.5402,
+      "step": 7000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5976e-05,
+      "loss": 12.3042,
+      "step": 8000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7976000000000002e-05,
+      "loss": 11.2017,
+      "step": 9000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9974000000000002e-05,
+      "loss": 10.2353,
+      "step": 10000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1974e-05,
+      "loss": 9.5083,
+      "step": 11000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3974e-05,
+      "loss": 9.0682,
+      "step": 12000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5974000000000002e-05,
+      "loss": 8.6615,
+      "step": 13000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7974e-05,
+      "loss": 8.4494,
+      "step": 14000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9974e-05,
+      "loss": 8.1137,
+      "step": 15000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1974e-05,
+      "loss": 7.8827,
+      "step": 16000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3974e-05,
+      "loss": 7.6477,
+      "step": 17000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5972e-05,
+      "loss": 7.4429,
+      "step": 18000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7972e-05,
+      "loss": 7.5768,
+      "step": 19000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9972e-05,
+      "loss": 7.467,
+      "step": 20000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1972e-05,
+      "loss": 7.4095,
+      "step": 21000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972000000000005e-05,
+      "loss": 7.4893,
+      "step": 22000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.597e-05,
+      "loss": 7.426,
+      "step": 23000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7968e-05,
+      "loss": 7.2462,
+      "step": 24000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.996800000000001e-05,
+      "loss": 7.2831,
+      "step": 25000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.196800000000001e-05,
+      "loss": 7.0669,
+      "step": 26000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.396800000000001e-05,
+      "loss": 6.8339,
+      "step": 27000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.5968e-05,
+      "loss": 6.7626,
+      "step": 28000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.7968e-05,
+      "loss": 6.8304,
+      "step": 29000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9968e-05,
+      "loss": 6.2205,
+      "step": 30000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.1968e-05,
+      "loss": 5.5114,
+      "step": 31000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.396800000000001e-05,
+      "loss": 4.9834,
+      "step": 32000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.5968e-05,
+      "loss": 4.4191,
+      "step": 33000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.7968e-05,
+      "loss": 3.2657,
+      "step": 34000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 6.9968e-05,
+      "loss": 3.1636,
+      "step": 35000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.1968e-05,
+      "loss": 2.9531,
+      "step": 36000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.3966e-05,
+      "loss": 2.8223,
+      "step": 37000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.5966e-05,
+      "loss": 2.8884,
+      "step": 38000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.7966e-05,
+      "loss": 2.9467,
+      "step": 39000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.996600000000001e-05,
+      "loss": 2.6816,
+      "step": 40000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.196600000000001e-05,
+      "loss": 2.4951,
+      "step": 41000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.3966e-05,
+      "loss": 2.3872,
+      "step": 42000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.5966e-05,
+      "loss": 2.382,
+      "step": 43000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.7966e-05,
+      "loss": 2.0369,
+      "step": 44000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.996600000000001e-05,
+      "loss": 2.0121,
+      "step": 45000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.196600000000001e-05,
+      "loss": 1.8462,
+      "step": 46000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.396600000000001e-05,
+      "loss": 1.8439,
+      "step": 47000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.5964e-05,
+      "loss": 2.1532,
+      "step": 48000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.7964e-05,
+      "loss": 2.0483,
+      "step": 49000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.996400000000001e-05,
+      "loss": 2.1008,
+      "step": 50000
+    },
+    {
+      "epoch": 0.05,
+      "eval_accuracy": 0.6533359021520881,
+      "eval_loss": 2.052328586578369,
+      "eval_runtime": 12.2044,
+      "eval_samples_per_second": 409.687,
+      "eval_steps_per_second": 0.819,
+      "step": 50000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.99997363576723e-05,
+      "loss": 2.0612,
+      "step": 51000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.999892709828038e-05,
+      "loss": 2.0449,
+      "step": 52000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.999757052195638e-05,
+      "loss": 1.8157,
+      "step": 53000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.999566718021146e-05,
+      "loss": 1.6483,
+      "step": 54000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.999322253966728e-05,
+      "loss": 1.7841,
+      "step": 55000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.999022682890849e-05,
+      "loss": 1.7667,
+      "step": 56000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.998668443303824e-05,
+      "loss": 1.7684,
+      "step": 57000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.998259539079562e-05,
+      "loss": 1.9884,
+      "step": 58000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.997796465555275e-05,
+      "loss": 1.9409,
+      "step": 59000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.99727884618501e-05,
+      "loss": 1.9498,
+      "step": 60000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.996706086563065e-05,
+      "loss": 1.961,
+      "step": 61000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.996079338458734e-05,
+      "loss": 1.9519,
+      "step": 62000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.995397353976218e-05,
+      "loss": 1.6914,
+      "step": 63000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 9.994660740628572e-05,
+      "loss": 1.7069,
+      "step": 64000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.993869506471283e-05,
+      "loss": 1.6265,
+      "step": 65000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.993023660157159e-05,
+      "loss": 1.5434,
+      "step": 66000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.992123210936239e-05,
+      "loss": 1.5563,
+      "step": 67000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.991168168655695e-05,
+      "loss": 1.7248,
+      "step": 68000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.990158543759713e-05,
+      "loss": 1.7096,
+      "step": 69000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.98909543874054e-05,
+      "loss": 1.7824,
+      "step": 70000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.987976736887685e-05,
+      "loss": 1.7253,
+      "step": 71000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.98680468781231e-05,
+      "loss": 1.7548,
+      "step": 72000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.9855769578893e-05,
+      "loss": 1.418,
+      "step": 73000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.984294706495438e-05,
+      "loss": 1.3509,
+      "step": 74000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 9.982957947653222e-05,
+      "loss": 1.2693,
+      "step": 75000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.981568114447057e-05,
+      "loss": 1.2466,
+      "step": 76000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.980123912510795e-05,
+      "loss": 1.4623,
+      "step": 77000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.978623830325964e-05,
+      "loss": 1.5242,
+      "step": 78000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.977069302708569e-05,
+      "loss": 1.4762,
+      "step": 79000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.975461982795923e-05,
+      "loss": 1.6135,
+      "step": 80000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.97379867031067e-05,
+      "loss": 1.6171,
+      "step": 81000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.972080965160031e-05,
+      "loss": 1.5516,
+      "step": 82000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.970308886128555e-05,
+      "loss": 1.3369,
+      "step": 83000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 9.968488013163436e-05,
+      "loss": 1.3425,
+      "step": 84000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.966607408075336e-05,
+      "loss": 1.2589,
+      "step": 85000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.964672488964346e-05,
+      "loss": 1.1773,
+      "step": 86000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.962683276990424e-05,
+      "loss": 1.4146,
+      "step": 87000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.960639793907279e-05,
+      "loss": 1.5667,
+      "step": 88000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.958544186883626e-05,
+      "loss": 1.6244,
+      "step": 89000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.956392283431062e-05,
+      "loss": 1.6549,
+      "step": 90000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.954188410838409e-05,
+      "loss": 1.6582,
+      "step": 91000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.951928181053549e-05,
+      "loss": 1.5158,
+      "step": 92000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.949613797775468e-05,
+      "loss": 1.4381,
+      "step": 93000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.947245286313893e-05,
+      "loss": 1.3927,
+      "step": 94000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.944822672570486e-05,
+      "loss": 1.3224,
+      "step": 95000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.942348486729986e-05,
+      "loss": 1.3627,
+      "step": 96000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.939820360201583e-05,
+      "loss": 1.5653,
+      "step": 97000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.937235708951778e-05,
+      "loss": 1.5405,
+      "step": 98000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.934597064883276e-05,
+      "loss": 1.5795,
+      "step": 99000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.931904456851861e-05,
+      "loss": 1.5248,
+      "step": 100000
+    },
+    {
+      "epoch": 0.1,
+      "eval_accuracy": 0.76607755765641,
+      "eval_loss": 1.1575313806533813,
+      "eval_runtime": 11.3109,
+      "eval_samples_per_second": 442.05,
+      "eval_steps_per_second": 0.884,
+      "step": 100000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.929160687776363e-05,
+      "loss": 1.5151,
+      "step": 101000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.926363121944323e-05,
+      "loss": 1.3509,
+      "step": 102000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.923508908775465e-05,
+      "loss": 1.361,
+      "step": 103000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.920600852902e-05,
+      "loss": 1.2428,
+      "step": 104000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.917641974860546e-05,
+      "loss": 1.2109,
+      "step": 105000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.914626383334564e-05,
+      "loss": 1.2223,
+      "step": 106000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.911557046241845e-05,
+      "loss": 1.3193,
+      "step": 107000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.90843399714817e-05,
+      "loss": 1.2732,
+      "step": 108000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.905260473734215e-05,
+      "loss": 1.3843,
+      "step": 109000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.902033414410157e-05,
+      "loss": 1.4177,
+      "step": 110000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.898749543760183e-05,
+      "loss": 1.3797,
+      "step": 111000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.89541210117005e-05,
+      "loss": 1.1533,
+      "step": 112000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.89202112313749e-05,
+      "loss": 1.246,
+      "step": 113000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.888580117932034e-05,
+      "loss": 1.1726,
+      "step": 114000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.885082234290936e-05,
+      "loss": 1.0864,
+      "step": 115000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.881534506150948e-05,
+      "loss": 1.3172,
+      "step": 116000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.877929869757103e-05,
+      "loss": 1.3628,
+      "step": 117000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.874271889103715e-05,
+      "loss": 1.3292,
+      "step": 118000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.8705643420908e-05,
+      "loss": 1.3346,
+      "step": 119000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.86679984675373e-05,
+      "loss": 1.3594,
+      "step": 120000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.862985973161956e-05,
+      "loss": 1.4662,
+      "step": 121000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.859115127647795e-05,
+      "loss": 1.2652,
+      "step": 122000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.855195094141563e-05,
+      "loss": 1.3002,
+      "step": 123000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.851218067604352e-05,
+      "loss": 1.234,
+      "step": 124000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.847187988923588e-05,
+      "loss": 1.2216,
+      "step": 125000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.843104902171565e-05,
+      "loss": 1.4338,
+      "step": 126000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.83897301449066e-05,
+      "loss": 1.4297,
+      "step": 127000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.834784099026668e-05,
+      "loss": 1.4427,
+      "step": 128000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.830542311138371e-05,
+      "loss": 1.4926,
+      "step": 129000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.826252018198228e-05,
+      "loss": 1.4958,
+      "step": 130000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.821904677956975e-05,
+      "loss": 1.3699,
+      "step": 131000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.817513458876564e-05,
+      "loss": 1.2653,
+      "step": 132000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.813060808918262e-05,
+      "loss": 1.1692,
+      "step": 133000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.808555524098074e-05,
+      "loss": 1.1308,
+      "step": 134000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.803997653685072e-05,
+      "loss": 1.2077,
+      "step": 135000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.799391884154434e-05,
+      "loss": 1.3723,
+      "step": 136000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.794733734161261e-05,
+      "loss": 1.2926,
+      "step": 137000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.790018513149605e-05,
+      "loss": 1.3797,
+      "step": 138000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.785250909262922e-05,
+      "loss": 1.3112,
+      "step": 139000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.780430974638956e-05,
+      "loss": 1.2632,
+      "step": 140000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.775563660295586e-05,
+      "loss": 1.0677,
+      "step": 141000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.770639275096809e-05,
+      "loss": 1.1267,
+      "step": 142000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.765662718951243e-05,
+      "loss": 1.0702,
+      "step": 143000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.760634046281708e-05,
+      "loss": 0.9839,
+      "step": 144000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 9.755558418801454e-05,
+      "loss": 1.1214,
+      "step": 145000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.750425730609625e-05,
+      "loss": 1.3301,
+      "step": 146000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.74524109252309e-05,
+      "loss": 1.2632,
+      "step": 147000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.740009823673144e-05,
+      "loss": 1.3007,
+      "step": 148000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.734721508266946e-05,
+      "loss": 1.3467,
+      "step": 149000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.729386780642266e-05,
+      "loss": 1.3152,
+      "step": 150000
+    },
+    {
+      "epoch": 0.15,
+      "eval_accuracy": 0.767424813667675,
+      "eval_loss": 1.1280603408813477,
+      "eval_runtime": 11.4235,
+      "eval_samples_per_second": 437.694,
+      "eval_steps_per_second": 0.875,
+      "step": 150000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.723995019013185e-05,
+      "loss": 1.086,
+      "step": 151000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.718557065738742e-05,
+      "loss": 1.0641,
+      "step": 152000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.713062093504106e-05,
+      "loss": 1.0293,
+      "step": 153000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.707515579977516e-05,
+      "loss": 1.1034,
+      "step": 154000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.701917585814753e-05,
+      "loss": 1.338,
+      "step": 155000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.696268172234585e-05,
+      "loss": 1.3888,
+      "step": 156000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.690573127421747e-05,
+      "loss": 1.3721,
+      "step": 157000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.684826889792009e-05,
+      "loss": 1.4046,
+      "step": 158000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.67902369329362e-05,
+      "loss": 1.42,
+      "step": 159000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.673169327741768e-05,
+      "loss": 1.3545,
+      "step": 160000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.667263857158856e-05,
+      "loss": 1.2049,
+      "step": 161000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.661313328110268e-05,
+      "loss": 1.208,
+      "step": 162000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.655311925585616e-05,
+      "loss": 1.1505,
+      "step": 163000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 9.649253631382465e-05,
+      "loss": 1.1549,
+      "step": 164000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.643150628198306e-05,
+      "loss": 1.3294,
+      "step": 165000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.63699076456116e-05,
+      "loss": 1.2572,
+      "step": 166000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.630780191535927e-05,
+      "loss": 1.3036,
+      "step": 167000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.624518977040433e-05,
+      "loss": 1.3684,
+      "step": 168000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.6182135265721e-05,
+      "loss": 1.3411,
+      "step": 169000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.611851285573377e-05,
+      "loss": 1.1562,
+      "step": 170000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.605445047952064e-05,
+      "loss": 1.1598,
+      "step": 171000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.598982058476909e-05,
+      "loss": 1.0123,
+      "step": 172000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.592468775270784e-05,
+      "loss": 0.9223,
+      "step": 173000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.585905269561911e-05,
+      "loss": 0.9687,
+      "step": 174000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.579291613127737e-05,
+      "loss": 1.2193,
+      "step": 175000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.572634567018936e-05,
+      "loss": 1.1626,
+      "step": 176000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.565920876628414e-05,
+      "loss": 1.1968,
+      "step": 177000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.559164042597832e-05,
+      "loss": 1.2435,
+      "step": 178000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.552350611635926e-05,
+      "loss": 1.2996,
+      "step": 179000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.545487396897081e-05,
+      "loss": 1.0764,
+      "step": 180000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.53859528647097e-05,
+      "loss": 1.0733,
+      "step": 181000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.531632878672679e-05,
+      "loss": 1.0556,
+      "step": 182000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.52462091366326e-05,
+      "loss": 0.9726,
+      "step": 183000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.517559468124436e-05,
+      "loss": 1.1002,
+      "step": 184000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.510448619279047e-05,
+      "loss": 1.1708,
+      "step": 185000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.503295629676674e-05,
+      "loss": 1.1752,
+      "step": 186000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.496086257254829e-05,
+      "loss": 1.3322,
+      "step": 187000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.48882771635401e-05,
+      "loss": 1.3658,
+      "step": 188000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.48152008635246e-05,
+      "loss": 1.3593,
+      "step": 189000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.474170828257791e-05,
+      "loss": 1.1516,
+      "step": 190000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.466765309224318e-05,
+      "loss": 1.1632,
+      "step": 191000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.459310942361059e-05,
+      "loss": 1.1145,
+      "step": 192000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.45181533665246e-05,
+      "loss": 1.074,
+      "step": 193000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9.444263567865316e-05,
+      "loss": 1.2866,
+      "step": 194000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.436670821943329e-05,
+      "loss": 1.3152,
+      "step": 195000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.429029654288884e-05,
+      "loss": 1.3184,
+      "step": 196000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.421332426904754e-05,
+      "loss": 1.3529,
+      "step": 197000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.413586848537565e-05,
+      "loss": 1.3308,
+      "step": 198000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.405793003891719e-05,
+      "loss": 1.1764,
+      "step": 199000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.39795097819945e-05,
+      "loss": 1.1239,
+      "step": 200000
+    },
+    {
+      "epoch": 0.2,
+      "eval_accuracy": 0.7971212829130808,
+      "eval_loss": 0.9457727670669556,
+      "eval_runtime": 11.3916,
+      "eval_samples_per_second": 438.92,
+      "eval_steps_per_second": 0.878,
+      "step": 200000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.3900608572199e-05,
+      "loss": 1.114,
+      "step": 201000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.382122727238165e-05,
+      "loss": 1.0155,
+      "step": 202000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.374136675064364e-05,
+      "loss": 1.0197,
+      "step": 203000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.366110845784027e-05,
+      "loss": 1.2245,
+      "step": 204000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.358029259454708e-05,
+      "loss": 1.1102,
+      "step": 205000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.349900014415627e-05,
+      "loss": 1.115,
+      "step": 206000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.341731400113154e-05,
+      "loss": 1.1427,
+      "step": 207000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.333507152310676e-05,
+      "loss": 1.2215,
+      "step": 208000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.325235513968268e-05,
+      "loss": 1.0317,
+      "step": 209000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.316924918077827e-05,
+      "loss": 1.0061,
+      "step": 210000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.30855881770826e-05,
+      "loss": 0.9843,
+      "step": 211000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.300145599629752e-05,
+      "loss": 0.9863,
+      "step": 212000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.291685355847915e-05,
+      "loss": 1.0236,
+      "step": 213000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.283186709471781e-05,
+      "loss": 1.1892,
+      "step": 214000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 9.274641316485581e-05,
+      "loss": 1.17,
+      "step": 215000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.266049270153557e-05,
+      "loss": 1.1997,
+      "step": 216000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.257401993701012e-05,
+      "loss": 1.1973,
+      "step": 217000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.248708158982357e-05,
+      "loss": 1.1667,
+      "step": 218000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.23996786107198e-05,
+      "loss": 1.0233,
+      "step": 219000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.231190005346579e-05,
+      "loss": 1.118,
+      "step": 220000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.222374826427257e-05,
+      "loss": 1.1016,
+      "step": 221000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.21349585284373e-05,
+      "loss": 1.0464,
+      "step": 222000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.204570801144619e-05,
+      "loss": 1.2016,
+      "step": 223000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.195599768932858e-05,
+      "loss": 1.2954,
+      "step": 224000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.18660093386864e-05,
+      "loss": 1.2599,
+      "step": 225000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.177538326919512e-05,
+      "loss": 1.3015,
+      "step": 226000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.16843003508054e-05,
+      "loss": 1.3161,
+      "step": 227000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.159276157958547e-05,
+      "loss": 1.308,
+      "step": 228000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.150076795658864e-05,
+      "loss": 1.1081,
+      "step": 229000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.140832048784247e-05,
+      "loss": 1.1327,
+      "step": 230000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.131542018433763e-05,
+      "loss": 1.0167,
+      "step": 231000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.122206806201698e-05,
+      "loss": 0.9574,
+      "step": 232000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.112835916951715e-05,
+      "loss": 1.2086,
+      "step": 233000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.103410692640435e-05,
+      "loss": 1.231,
+      "step": 234000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.093940594087245e-05,
+      "loss": 1.181,
+      "step": 235000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.084435262053219e-05,
+      "loss": 1.2146,
+      "step": 236000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.074875770810744e-05,
+      "loss": 1.2343,
+      "step": 237000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.065271717379691e-05,
+      "loss": 1.0341,
+      "step": 238000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.055642548107829e-05,
+      "loss": 0.9415,
+      "step": 239000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.045949774468634e-05,
+      "loss": 0.9071,
+      "step": 240000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.036212754970922e-05,
+      "loss": 0.9272,
+      "step": 241000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.026431596097182e-05,
+      "loss": 0.932,
+      "step": 242000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.01661625196235e-05,
+      "loss": 1.0876,
+      "step": 243000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.006747179584774e-05,
+      "loss": 1.1039,
+      "step": 244000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.99683429006198e-05,
+      "loss": 1.1908,
+      "step": 245000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.986887670194317e-05,
+      "loss": 1.1791,
+      "step": 246000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.976887515621704e-05,
+      "loss": 1.1991,
+      "step": 247000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.966853935776588e-05,
+      "loss": 1.0425,
+      "step": 248000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.956777061770676e-05,
+      "loss": 1.0149,
+      "step": 249000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.946646851795914e-05,
+      "loss": 0.9472,
+      "step": 250000
+    },
+    {
+      "epoch": 0.25,
+      "eval_accuracy": 0.7875509927594045,
+      "eval_loss": 0.9979193806648254,
+      "eval_runtime": 10.7394,
+      "eval_samples_per_second": 465.574,
+      "eval_steps_per_second": 0.931,
+      "step": 250000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.936473481923489e-05,
+      "loss": 0.8647,
+      "step": 251000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.926257063407744e-05,
+      "loss": 1.0154,
+      "step": 252000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.916007988738915e-05,
+      "loss": 1.2556,
+      "step": 253000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.905705851349957e-05,
+      "loss": 1.2242,
+      "step": 254000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.895361001787558e-05,
+      "loss": 1.2757,
+      "step": 255000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.884973553181336e-05,
+      "loss": 1.2906,
+      "step": 256000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.874554070244375e-05,
+      "loss": 1.2885,
+      "step": 257000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.864092300505416e-05,
+      "loss": 1.0966,
+      "step": 258000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.853577822596901e-05,
+      "loss": 1.0932,
+      "step": 259000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.843021202578716e-05,
+      "loss": 1.0595,
+      "step": 260000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.83243317549683e-05,
+      "loss": 1.0109,
+      "step": 261000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.821792659908128e-05,
+      "loss": 1.1995,
+      "step": 262000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.811110349807205e-05,
+      "loss": 1.2683,
+      "step": 263000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.800397106780893e-05,
+      "loss": 1.1763,
+      "step": 264000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.789631600073084e-05,
+      "loss": 1.1731,
+      "step": 265000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.778835478172145e-05,
+      "loss": 1.2495,
+      "step": 266000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.76798724530479e-05,
+      "loss": 1.2205,
+      "step": 267000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.757097806332607e-05,
+      "loss": 0.9901,
+      "step": 268000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.746167280340752e-05,
+      "loss": 1.0082,
+      "step": 269000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.735217770656539e-05,
+      "loss": 0.9372,
+      "step": 270000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.724205511251769e-05,
+      "loss": 0.8586,
+      "step": 271000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.713152524532197e-05,
+      "loss": 1.0177,
+      "step": 272000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.702058931371517e-05,
+      "loss": 1.0323,
+      "step": 273000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.6909360073476e-05,
+      "loss": 1.11,
+      "step": 274000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.67976160600311e-05,
+      "loss": 1.1162,
+      "step": 275000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.668558198077614e-05,
+      "loss": 1.1175,
+      "step": 276000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.65730347686515e-05,
+      "loss": 1.0497,
+      "step": 277000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.646008759967813e-05,
+      "loss": 1.0092,
+      "step": 278000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.634674170902821e-05,
+      "loss": 0.9418,
+      "step": 279000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.623311227773569e-05,
+      "loss": 0.9255,
+      "step": 280000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.611897306229244e-05,
+      "loss": 0.9643,
+      "step": 281000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.60045535866349e-05,
+      "loss": 1.1115,
+      "step": 282000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.58896260342258e-05,
+      "loss": 1.0597,
+      "step": 283000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.577442151427801e-05,
+      "loss": 1.0903,
+      "step": 284000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 8.565871064718052e-05,
+      "loss": 1.1758,
+      "step": 285000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.554272611732321e-05,
+      "loss": 1.2468,
+      "step": 286000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.542623699204519e-05,
+      "loss": 1.0986,
+      "step": 287000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.53093604511152e-05,
+      "loss": 1.0916,
+      "step": 288000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.519209777267638e-05,
+      "loss": 1.0491,
+      "step": 289000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.507456807843545e-05,
+      "loss": 1.0027,
+      "step": 290000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.495653735920981e-05,
+      "loss": 1.106,
+      "step": 291000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.483824296440821e-05,
+      "loss": 1.2315,
+      "step": 292000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.471944936228e-05,
+      "loss": 1.1994,
+      "step": 293000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.460051479845447e-05,
+      "loss": 1.2447,
+      "step": 294000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 8.448096388237765e-05,
+      "loss": 1.2671,
+      "step": 295000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.436103588800417e-05,
+      "loss": 1.2646,
+      "step": 296000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.424085261786644e-05,
+      "loss": 0.9903,
+      "step": 297000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.412017477934155e-05,
+      "loss": 0.9825,
+      "step": 298000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.399912380805182e-05,
+      "loss": 0.9979,
+      "step": 299000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.387770102779126e-05,
+      "loss": 0.961,
+      "step": 300000
+    },
+    {
+      "epoch": 0.3,
+      "eval_accuracy": 0.8074508577716577,
+      "eval_loss": 0.8797607421875,
+      "eval_runtime": 10.8875,
+      "eval_samples_per_second": 459.241,
+      "eval_steps_per_second": 0.918,
+      "step": 300000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.375602974429366e-05,
+      "loss": 1.073,
+      "step": 301000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.363399004819421e-05,
+      "loss": 1.144,
+      "step": 302000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.351146055866368e-05,
+      "loss": 1.0968,
+      "step": 303000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.338856459316948e-05,
+      "loss": 1.0641,
+      "step": 304000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 8.326530349568222e-05,
+      "loss": 1.064,
+      "step": 305000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.31416786141655e-05,
+      "loss": 1.0356,
+      "step": 306000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.301781546845868e-05,
+      "loss": 0.9338,
+      "step": 307000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.289346743906984e-05,
+      "loss": 0.9026,
+      "step": 308000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.276888457896356e-05,
+      "loss": 0.8446,
+      "step": 309000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.264381883565022e-05,
+      "loss": 0.9099,
+      "step": 310000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.251839610476409e-05,
+      "loss": 1.1068,
+      "step": 311000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.239261775790807e-05,
+      "loss": 1.0591,
+      "step": 312000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.226648517057399e-05,
+      "loss": 1.1183,
+      "step": 313000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 8.213999972212765e-05,
+      "loss": 1.146,
+      "step": 314000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.201328980782183e-05,
+      "loss": 1.1369,
+      "step": 315000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.188610314006513e-05,
+      "loss": 0.9698,
+      "step": 316000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.175856777099312e-05,
+      "loss": 0.9333,
+      "step": 317000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.163068509531211e-05,
+      "loss": 0.9574,
+      "step": 318000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.150271331297903e-05,
+      "loss": 0.9808,
+      "step": 319000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.137414091098588e-05,
+      "loss": 1.0482,
+      "step": 320000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.124535449282972e-05,
+      "loss": 1.2092,
+      "step": 321000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.111622706352883e-05,
+      "loss": 1.1759,
+      "step": 322000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.098663026608033e-05,
+      "loss": 1.2193,
+      "step": 323000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 8.085669460380775e-05,
+      "loss": 1.2402,
+      "step": 324000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.072642149766666e-05,
+      "loss": 1.2365,
+      "step": 325000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.059581237230287e-05,
+      "loss": 1.0715,
+      "step": 326000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.046486865603688e-05,
+      "loss": 1.0666,
+      "step": 327000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.033359178084814e-05,
+      "loss": 1.0308,
+      "step": 328000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.020211495617513e-05,
+      "loss": 0.9836,
+      "step": 329000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 8.007017640320121e-05,
+      "loss": 1.0282,
+      "step": 330000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.993804143876718e-05,
+      "loss": 1.1226,
+      "step": 331000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.980544697365782e-05,
+      "loss": 1.1466,
+      "step": 332000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.967265964369563e-05,
+      "loss": 1.1792,
+      "step": 333000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 7.953954846694738e-05,
+      "loss": 1.134,
+      "step": 334000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.940598116801423e-05,
+      "loss": 1.159,
+      "step": 335000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.927209228998228e-05,
+      "loss": 0.9188,
+      "step": 336000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.913801766544098e-05,
+      "loss": 0.8693,
+      "step": 337000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.900349034318708e-05,
+      "loss": 0.8056,
+      "step": 338000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.886864584340953e-05,
+      "loss": 0.808,
+      "step": 339000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.873348564074622e-05,
+      "loss": 1.0187,
+      "step": 340000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.859828247475456e-05,
+      "loss": 1.0268,
+      "step": 341000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.846249592803267e-05,
+      "loss": 1.0245,
+      "step": 342000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.832653437280335e-05,
+      "loss": 1.1168,
+      "step": 343000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.819012710086294e-05,
+      "loss": 1.1404,
+      "step": 344000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 7.8053411546209e-05,
+      "loss": 1.0336,
+      "step": 345000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.791638920394093e-05,
+      "loss": 0.9649,
+      "step": 346000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.777906157251316e-05,
+      "loss": 0.9322,
+      "step": 347000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.76414301537186e-05,
+      "loss": 0.8737,
+      "step": 348000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.750363453686169e-05,
+      "loss": 0.855,
+      "step": 349000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.736567712952333e-05,
+      "loss": 1.0179,
+      "step": 350000
+    },
+    {
+      "epoch": 0.35,
+      "eval_accuracy": 0.8018313508243486,
+      "eval_loss": 0.9101699590682983,
+      "eval_runtime": 11.7713,
+      "eval_samples_per_second": 424.764,
+      "eval_steps_per_second": 0.85,
+      "step": 350000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.722714428804807e-05,
+      "loss": 1.096,
+      "step": 351000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.708831369488505e-05,
+      "loss": 1.1769,
+      "step": 352000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.694918686826337e-05,
+      "loss": 1.218,
+      "step": 353000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.680976532965179e-05,
+      "loss": 1.2225,
+      "step": 354000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 7.66701904644065e-05,
+      "loss": 1.0765,
+      "step": 355000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.653018436999124e-05,
+      "loss": 1.0637,
+      "step": 356000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.638988814573021e-05,
+      "loss": 1.0175,
+      "step": 357000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.624930332588043e-05,
+      "loss": 0.9619,
+      "step": 358000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.61085724626061e-05,
+      "loss": 1.0432,
+      "step": 359000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.596755665091857e-05,
+      "loss": 1.1913,
+      "step": 360000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.582625742986147e-05,
+      "loss": 1.1628,
+      "step": 361000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.568453447806063e-05,
+      "loss": 1.2034,
+      "step": 362000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.554253064430575e-05,
+      "loss": 1.1454,
+      "step": 363000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.540024748152803e-05,
+      "loss": 1.1433,
+      "step": 364000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 7.525768654571332e-05,
+      "loss": 1.0084,
+      "step": 365000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.511484939588518e-05,
+      "loss": 0.9967,
+      "step": 366000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.497188084255762e-05,
+      "loss": 0.9075,
+      "step": 367000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.482849622614294e-05,
+      "loss": 0.8931,
+      "step": 368000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.468512767147887e-05,
+      "loss": 0.9603,
+      "step": 369000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.454120212348005e-05,
+      "loss": 0.9906,
+      "step": 370000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.43970081968267e-05,
+      "loss": 0.9656,
+      "step": 371000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.425254746840043e-05,
+      "loss": 1.0444,
+      "step": 372000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.410796637590287e-05,
+      "loss": 1.0772,
+      "step": 373000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 7.396297704907714e-05,
+      "loss": 1.0469,
+      "step": 374000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.381772566697359e-05,
+      "loss": 0.8572,
+      "step": 375000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.36722138180381e-05,
+      "loss": 0.936,
+      "step": 376000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.352658899306739e-05,
+      "loss": 0.8848,
+      "step": 377000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.338070739939699e-05,
+      "loss": 0.8367,
+      "step": 378000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.323442421881121e-05,
+      "loss": 1.056,
+      "step": 379000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.308788695028466e-05,
+      "loss": 1.0573,
+      "step": 380000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.294109719632544e-05,
+      "loss": 1.0499,
+      "step": 381000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.279420372761656e-05,
+      "loss": 1.0515,
+      "step": 382000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.26469140698113e-05,
+      "loss": 1.0599,
+      "step": 383000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 7.249952440947382e-05,
+      "loss": 1.0554,
+      "step": 384000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.235174128430801e-05,
+      "loss": 1.04,
+      "step": 385000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.220371372408057e-05,
+      "loss": 1.0006,
+      "step": 386000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.20554433475972e-05,
+      "loss": 0.9656,
+      "step": 387000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.190708040782669e-05,
+      "loss": 0.9891,
+      "step": 388000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.175832950461096e-05,
+      "loss": 1.1548,
+      "step": 389000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.16094897629503e-05,
+      "loss": 1.1437,
+      "step": 390000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.146041417577084e-05,
+      "loss": 1.1659,
+      "step": 391000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.131095479401597e-05,
+      "loss": 1.2044,
+      "step": 392000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.116126235907376e-05,
+      "loss": 1.2036,
+      "step": 393000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 7.101133850795673e-05,
+      "loss": 1.0658,
+      "step": 394000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.086118488020812e-05,
+      "loss": 1.0297,
+      "step": 395000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.071095361305229e-05,
+      "loss": 0.9179,
+      "step": 396000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.056034558637109e-05,
+      "loss": 0.8803,
+      "step": 397000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.040966365967704e-05,
+      "loss": 0.9978,
+      "step": 398000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.025860781555982e-05,
+      "loss": 1.1169,
+      "step": 399000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 7.010748181405352e-05,
+      "loss": 1.037,
+      "step": 400000
+    },
+    {
+      "epoch": 0.4,
+      "eval_accuracy": 0.8194723236238552,
+      "eval_loss": 0.810663104057312,
+      "eval_runtime": 10.4613,
+      "eval_samples_per_second": 477.953,
+      "eval_steps_per_second": 0.956,
+      "step": 400000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.995598475356628e-05,
+      "loss": 1.1173,
+      "step": 401000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.980442128137954e-05,
+      "loss": 1.0554,
+      "step": 402000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.965248962486922e-05,
+      "loss": 1.0148,
+      "step": 403000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 6.950049530525111e-05,
+      "loss": 0.8313,
+      "step": 404000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.934813569205621e-05,
+      "loss": 0.8747,
+      "step": 405000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.919571716708366e-05,
+      "loss": 0.8472,
+      "step": 406000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.904293625524425e-05,
+      "loss": 0.7775,
+      "step": 407000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.88899470929049e-05,
+      "loss": 0.9395,
+      "step": 408000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.873675135313058e-05,
+      "loss": 1.0758,
+      "step": 409000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.858350421367768e-05,
+      "loss": 1.041,
+      "step": 410000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.842990054963225e-05,
+      "loss": 1.0575,
+      "step": 411000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.827624924447044e-05,
+      "loss": 1.1056,
+      "step": 412000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.812224436855836e-05,
+      "loss": 1.075,
+      "step": 413000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.796819561217397e-05,
+      "loss": 0.8725,
+      "step": 414000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 6.781379625433246e-05,
+      "loss": 0.8658,
+      "step": 415000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.765920208766913e-05,
+      "loss": 0.8149,
+      "step": 416000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.75045696859852e-05,
+      "loss": 0.9049,
+      "step": 417000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.734959116622053e-05,
+      "loss": 1.1137,
+      "step": 418000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.719457817656532e-05,
+      "loss": 1.1364,
+      "step": 419000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.703922207618428e-05,
+      "loss": 1.146,
+      "step": 420000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.688383527254907e-05,
+      "loss": 1.1727,
+      "step": 421000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.672810838053671e-05,
+      "loss": 1.19,
+      "step": 422000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.657251055370414e-05,
+      "loss": 1.0855,
+      "step": 423000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.641641985562924e-05,
+      "loss": 1.019,
+      "step": 424000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 6.626014963021694e-05,
+      "loss": 0.9942,
+      "step": 425000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.61037015864134e-05,
+      "loss": 0.9554,
+      "step": 426000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.594707743510933e-05,
+      "loss": 0.9881,
+      "step": 427000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.57902788891214e-05,
+      "loss": 1.1244,
+      "step": 428000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.563346472008175e-05,
+      "loss": 1.0419,
+      "step": 429000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.547632270089105e-05,
+      "loss": 1.0964,
+      "step": 430000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.531916883034684e-05,
+      "loss": 1.1526,
+      "step": 431000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.516169020498223e-05,
+      "loss": 1.1175,
+      "step": 432000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.500420350048018e-05,
+      "loss": 0.945,
+      "step": 433000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.48465530404153e-05,
+      "loss": 0.9563,
+      "step": 434000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 6.468858249431153e-05,
+      "loss": 0.8274,
+      "step": 435000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.453045131622291e-05,
+      "loss": 0.7477,
+      "step": 436000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.437216123544669e-05,
+      "loss": 0.835,
+      "step": 437000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.421403103322858e-05,
+      "loss": 1.0134,
+      "step": 438000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.40554286510482e-05,
+      "loss": 0.973,
+      "step": 439000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.38968313932415e-05,
+      "loss": 1.0058,
+      "step": 440000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.373792348246463e-05,
+      "loss": 1.0381,
+      "step": 441000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.357886533595756e-05,
+      "loss": 1.0955,
+      "step": 442000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.341965869315472e-05,
+      "loss": 0.8979,
+      "step": 443000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 6.326030529511443e-05,
+      "loss": 0.8869,
+      "step": 444000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.310096645476388e-05,
+      "loss": 0.8761,
+      "step": 445000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.294132491822075e-05,
+      "loss": 0.8092,
+      "step": 446000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.278170171059243e-05,
+      "loss": 0.9651,
+      "step": 447000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.2621779011803e-05,
+      "loss": 0.9916,
+      "step": 448000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.24618784123524e-05,
+      "loss": 0.9917,
+      "step": 449000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.230168153985983e-05,
+      "loss": 1.1206,
+      "step": 450000
+    },
+    {
+      "epoch": 0.45,
+      "eval_accuracy": 0.8152173260941297,
+      "eval_loss": 0.8323125839233398,
+      "eval_runtime": 10.9195,
+      "eval_samples_per_second": 457.898,
+      "eval_steps_per_second": 0.916,
+      "step": 450000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.214167093404725e-05,
+      "loss": 1.1638,
+      "step": 451000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.198120702027724e-05,
+      "loss": 1.1351,
+      "step": 452000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.182061208194684e-05,
+      "loss": 0.9823,
+      "step": 453000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 6.165988787529652e-05,
+      "loss": 0.99,
+      "step": 454000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.149903615798047e-05,
+      "loss": 0.9387,
+      "step": 455000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.133821972874338e-05,
+      "loss": 0.9391,
+      "step": 456000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.117711839172833e-05,
+      "loss": 1.1115,
+      "step": 457000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.1016217391488694e-05,
+      "loss": 1.1276,
+      "step": 458000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.085487359441242e-05,
+      "loss": 1.1379,
+      "step": 459000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.069341109017761e-05,
+      "loss": 1.172,
+      "step": 460000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.053183164451236e-05,
+      "loss": 1.1493,
+      "step": 461000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.037013702442358e-05,
+      "loss": 0.9853,
+      "step": 462000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.020849086226153e-05,
+      "loss": 0.9542,
+      "step": 463000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 6.004657131011767e-05,
+      "loss": 0.9451,
+      "step": 464000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.988454189027957e-05,
+      "loss": 0.8593,
+      "step": 465000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.972256656559438e-05,
+      "loss": 0.8856,
+      "step": 466000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.956032283276984e-05,
+      "loss": 1.0615,
+      "step": 467000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.939813694970145e-05,
+      "loss": 0.9391,
+      "step": 468000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.923584849556318e-05,
+      "loss": 0.961,
+      "step": 469000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.9073296639853294e-05,
+      "loss": 0.9814,
+      "step": 470000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8910645560025016e-05,
+      "loss": 1.0385,
+      "step": 471000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.874789703480451e-05,
+      "loss": 0.8768,
+      "step": 472000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8585052843983566e-05,
+      "loss": 0.8435,
+      "step": 473000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.842211476840016e-05,
+      "loss": 0.8323,
+      "step": 474000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.825924766550955e-05,
+      "loss": 0.8381,
+      "step": 475000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.8096290421362876e-05,
+      "loss": 0.9095,
+      "step": 476000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.793308156198103e-05,
+      "loss": 1.0265,
+      "step": 477000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.776978594769021e-05,
+      "loss": 1.0133,
+      "step": 478000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7606405364265115e-05,
+      "loss": 1.0417,
+      "step": 479000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.744310510313018e-05,
+      "loss": 1.0367,
+      "step": 480000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.727956002295942e-05,
+      "loss": 1.0063,
+      "step": 481000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.711593533468653e-05,
+      "loss": 0.8627,
+      "step": 482000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.6952232827684906e-05,
+      "loss": 0.9525,
+      "step": 483000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.678861810809461e-05,
+      "loss": 0.937,
+      "step": 484000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.662476540848262e-05,
+      "loss": 0.9029,
+      "step": 485000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.646116818290258e-05,
+      "loss": 1.0793,
+      "step": 486000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.629717252072332e-05,
+      "loss": 1.1195,
+      "step": 487000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.613327209200223e-05,
+      "loss": 1.1135,
+      "step": 488000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.596914056044104e-05,
+      "loss": 1.14,
+      "step": 489000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.580494375131501e-05,
+      "loss": 1.1599,
+      "step": 490000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.5640683460254154e-05,
+      "loss": 1.1177,
+      "step": 491000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.547636148358269e-05,
+      "loss": 0.9853,
+      "step": 492000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.531197961829942e-05,
+      "loss": 0.982,
+      "step": 493000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.514770413043232e-05,
+      "loss": 0.8807,
+      "step": 494000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.498320793691619e-05,
+      "loss": 0.8529,
+      "step": 495000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.481882182514718e-05,
+      "loss": 1.0643,
+      "step": 496000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.465421849180081e-05,
+      "loss": 1.0695,
+      "step": 497000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.4489564260666694e-05,
+      "loss": 1.0373,
+      "step": 498000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.432502565962998e-05,
+      "loss": 1.0638,
+      "step": 499000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.4160275081750255e-05,
+      "loss": 1.0865,
+      "step": 500000
+    },
+    {
+      "epoch": 0.5,
+      "eval_accuracy": 0.8242477038501768,
+      "eval_loss": 0.782859206199646,
+      "eval_runtime": 10.6001,
+      "eval_samples_per_second": 471.696,
+      "eval_steps_per_second": 0.943,
+      "step": 500000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.39956438259683e-05,
+      "loss": 0.8685,
+      "step": 501000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.3830804100862085e-05,
+      "loss": 0.8195,
+      "step": 502000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.3666252286543196e-05,
+      "loss": 0.7759,
+      "step": 503000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.350133065681624e-05,
+      "loss": 0.7921,
+      "step": 504000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.3336370737098185e-05,
+      "loss": 0.8275,
+      "step": 505000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.3171374331364276e-05,
+      "loss": 0.9585,
+      "step": 506000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.300650829179897e-05,
+      "loss": 0.9638,
+      "step": 507000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.284144435951072e-05,
+      "loss": 1.0516,
+      "step": 508000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.2676349353642174e-05,
+      "loss": 1.0423,
+      "step": 509000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.251139021793834e-05,
+      "loss": 1.0497,
+      "step": 510000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.234623850814717e-05,
+      "loss": 0.9067,
+      "step": 511000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.218122632985052e-05,
+      "loss": 0.8785,
+      "step": 512000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.201602513318512e-05,
+      "loss": 0.8185,
+      "step": 513000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.185080188959187e-05,
+      "loss": 0.752,
+      "step": 514000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.168572365891783e-05,
+      "loss": 0.9095,
+      "step": 515000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.152046175978524e-05,
+      "loss": 1.0988,
+      "step": 516000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.1355348519357e-05,
+      "loss": 1.0842,
+      "step": 517000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.11900551865487e-05,
+      "loss": 1.125,
+      "step": 518000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.102491415173354e-05,
+      "loss": 1.1397,
+      "step": 519000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.08595966084867e-05,
+      "loss": 1.1343,
+      "step": 520000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.069443499586649e-05,
+      "loss": 0.9646,
+      "step": 521000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.052910046647634e-05,
+      "loss": 0.9628,
+      "step": 522000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.036376015092827e-05,
+      "loss": 0.9286,
+      "step": 523000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.0198415857357464e-05,
+      "loss": 0.9084,
+      "step": 524000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.003323474088806e-05,
+      "loss": 1.0839,
+      "step": 525000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.986805326169539e-05,
+      "loss": 1.1188,
+      "step": 526000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.970270787850604e-05,
+      "loss": 1.0571,
+      "step": 527000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.9537365746455695e-05,
+      "loss": 1.0487,
+      "step": 528000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.937202867369945e-05,
+      "loss": 1.1107,
+      "step": 529000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.920686379451033e-05,
+      "loss": 1.0362,
+      "step": 530000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.904170757159314e-05,
+      "loss": 0.9012,
+      "step": 531000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.887639650222643e-05,
+      "loss": 0.8745,
+      "step": 532000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8711097720407546e-05,
+      "loss": 0.8339,
+      "step": 533000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.854581303381751e-05,
+      "loss": 0.7654,
+      "step": 534000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.838070951022198e-05,
+      "loss": 0.9112,
+      "step": 535000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.821545841788341e-05,
+      "loss": 0.8959,
+      "step": 536000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.805039206223316e-05,
+      "loss": 0.986,
+      "step": 537000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.78851817855374e-05,
+      "loss": 0.9942,
+      "step": 538000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.772015981115084e-05,
+      "loss": 0.9933,
+      "step": 539000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.7554997569705646e-05,
+      "loss": 0.8891,
+      "step": 540000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.7389862066415265e-05,
+      "loss": 0.9105,
+      "step": 541000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.722475510717516e-05,
+      "loss": 0.8341,
+      "step": 542000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.7059843558417655e-05,
+      "loss": 0.8087,
+      "step": 543000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.689479907063946e-05,
+      "loss": 0.8851,
+      "step": 544000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.672995353380826e-05,
+      "loss": 0.9986,
+      "step": 545000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.656497872985476e-05,
+      "loss": 0.9483,
+      "step": 546000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.6400371326533076e-05,
+      "loss": 0.9711,
+      "step": 547000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.623547337545436e-05,
+      "loss": 1.0405,
+      "step": 548000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.607061659263568e-05,
+      "loss": 1.113,
+      "step": 549000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.5905802780924407e-05,
+      "loss": 0.9616,
+      "step": 550000
+    },
+    {
+      "epoch": 0.55,
+      "eval_accuracy": 0.822435579595823,
+      "eval_loss": 0.7894766926765442,
+      "eval_runtime": 10.8513,
+      "eval_samples_per_second": 460.775,
+      "eval_steps_per_second": 0.922,
+      "step": 550000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.574103374269796e-05,
+      "loss": 0.9594,
+      "step": 551000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.557647597844301e-05,
+      "loss": 0.9291,
+      "step": 552000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.541180184306395e-05,
+      "loss": 0.8842,
+      "step": 553000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.5247507080131874e-05,
+      "loss": 1.0254,
+      "step": 554000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.508293499090543e-05,
+      "loss": 1.1074,
+      "step": 555000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4918416673914135e-05,
+      "loss": 1.0816,
+      "step": 556000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4754118362693004e-05,
+      "loss": 1.1233,
+      "step": 557000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.458971292873412e-05,
+      "loss": 1.1443,
+      "step": 558000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.442536666080782e-05,
+      "loss": 1.1356,
+      "step": 559000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.426124561043402e-05,
+      "loss": 0.88,
+      "step": 560000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.409702300204254e-05,
+      "loss": 0.8733,
+      "step": 561000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.393286494766179e-05,
+      "loss": 0.8834,
+      "step": 562000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.376877324249791e-05,
+      "loss": 0.8683,
+      "step": 563000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.3604913669958144e-05,
+      "loss": 0.9784,
+      "step": 564000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.344095997509148e-05,
+      "loss": 1.0091,
+      "step": 565000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.32770780088357e-05,
+      "loss": 0.994,
+      "step": 566000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.3113433334503214e-05,
+      "loss": 0.9519,
+      "step": 567000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.29497001250192e-05,
+      "loss": 0.9524,
+      "step": 568000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.278620763348594e-05,
+      "loss": 0.8762,
+      "step": 569000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2622630335834316e-05,
+      "loss": 0.8521,
+      "step": 570000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.245929717164161e-05,
+      "loss": 0.7913,
+      "step": 571000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.229588293406972e-05,
+      "loss": 0.7513,
+      "step": 572000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.2132552947487734e-05,
+      "loss": 0.8198,
+      "step": 573000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.196930899804627e-05,
+      "loss": 1.006,
+      "step": 574000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.1806479094376326e-05,
+      "loss": 0.9535,
+      "step": 575000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1643412392891206e-05,
+      "loss": 1.0172,
+      "step": 576000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.1480600006783755e-05,
+      "loss": 1.0338,
+      "step": 577000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.131771776611856e-05,
+      "loss": 1.0166,
+      "step": 578000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.115493047350083e-05,
+      "loss": 0.8639,
+      "step": 579000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.099223990914634e-05,
+      "loss": 0.8269,
+      "step": 580000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.082981039447362e-05,
+      "loss": 0.8496,
+      "step": 581000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0667318521868897e-05,
+      "loss": 0.8634,
+      "step": 582000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.0504928709973586e-05,
+      "loss": 0.9634,
+      "step": 583000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.034280496817482e-05,
+      "loss": 1.1009,
+      "step": 584000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.018062449767229e-05,
+      "loss": 1.0633,
+      "step": 585000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.001871342914655e-05,
+      "loss": 1.1073,
+      "step": 586000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.98567493872435e-05,
+      "loss": 1.1197,
+      "step": 587000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.9695219864658135e-05,
+      "loss": 1.121,
+      "step": 588000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.953364089929742e-05,
+      "loss": 0.9561,
+      "step": 589000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.937201459501688e-05,
+      "loss": 0.9589,
+      "step": 590000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.921050451668199e-05,
+      "loss": 0.9277,
+      "step": 591000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.9049112430541065e-05,
+      "loss": 0.8929,
+      "step": 592000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.888800131347529e-05,
+      "loss": 0.9633,
+      "step": 593000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.872685038288536e-05,
+      "loss": 1.0104,
+      "step": 594000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 3.856582273365353e-05,
+      "loss": 1.0472,
+      "step": 595000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.840492012675236e-05,
+      "loss": 1.0702,
+      "step": 596000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.8244305033669075e-05,
+      "loss": 1.0294,
+      "step": 597000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.808365765942003e-05,
+      "loss": 1.0254,
+      "step": 598000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.7923140600381655e-05,
+      "loss": 0.8297,
+      "step": 599000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.776275561194279e-05,
+      "loss": 0.7727,
+      "step": 600000
+    },
+    {
+      "epoch": 0.6,
+      "eval_accuracy": 0.8285096268267774,
+      "eval_loss": 0.7584890127182007,
+      "eval_runtime": 11.327,
+      "eval_samples_per_second": 441.425,
+      "eval_steps_per_second": 0.883,
+      "step": 600000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.760250444804797e-05,
+      "loss": 0.7138,
+      "step": 601000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.744254890846151e-05,
+      "loss": 0.7417,
+      "step": 602000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.7282730420633285e-05,
+      "loss": 0.929,
+      "step": 603000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.712289095940718e-05,
+      "loss": 0.9185,
+      "step": 604000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 3.696319232018241e-05,
+      "loss": 0.9308,
+      "step": 605000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.6803636249397785e-05,
+      "loss": 1.0135,
+      "step": 606000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.664438383102547e-05,
+      "loss": 1.0432,
+      "step": 607000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.648511798325486e-05,
+      "loss": 0.9097,
+      "step": 608000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.632599993206869e-05,
+      "loss": 0.8698,
+      "step": 609000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 3.616719031079871e-05,
+      "loss": 0.8344,
+      "step": 610000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600837291927178e-05,
+      "loss": 0.8786,
+      "step": 611000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.584970853793694e-05,
+      "loss": 0.9996,
+      "step": 612000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.569135733368609e-05,
+      "loss": 1.0586,
+      "step": 613000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.553300401908391e-05,
+      "loss": 1.0897,
+      "step": 614000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.5374967028736584e-05,
+      "loss": 0.9494,
+      "step": 615000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.521693170082889e-05,
+      "loss": 0.856,
+      "step": 616000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.505905803818765e-05,
+      "loss": 0.8594,
+      "step": 617000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.490150539537649e-05,
+      "loss": 0.7276,
+      "step": 618000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.474396007494727e-05,
+      "loss": 0.7412,
+      "step": 619000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.45867388867009e-05,
+      "loss": 0.8891,
+      "step": 620000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.442952879313661e-05,
+      "loss": 0.9131,
+      "step": 621000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.427264592994348e-05,
+      "loss": 0.9492,
+      "step": 622000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.411577793324775e-05,
+      "loss": 0.9621,
+      "step": 623000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.395924025081174e-05,
+      "loss": 0.975,
+      "step": 624000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3802721206038737e-05,
+      "loss": 0.8927,
+      "step": 625000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.36463792921116e-05,
+      "loss": 0.8459,
+      "step": 626000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3490372291934334e-05,
+      "loss": 0.8157,
+      "step": 627000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.333438958553279e-05,
+      "loss": 0.7887,
+      "step": 628000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.3178744840428254e-05,
+      "loss": 0.8727,
+      "step": 629000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.302312815791971e-05,
+      "loss": 0.9689,
+      "step": 630000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2867697131761696e-05,
+      "loss": 0.9148,
+      "step": 631000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2712608611243435e-05,
+      "loss": 0.9737,
+      "step": 632000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255770876495283e-05,
+      "loss": 1.0775,
+      "step": 633000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.240284451505519e-05,
+      "loss": 1.1036,
+      "step": 634000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2248172704830614e-05,
+      "loss": 0.9381,
+      "step": 635000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.209369502574527e-05,
+      "loss": 0.9412,
+      "step": 636000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.1939413167142243e-05,
+      "loss": 0.9021,
+      "step": 637000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.178532881622327e-05,
+      "loss": 0.8694,
+      "step": 638000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.163159744313126e-05,
+      "loss": 1.0402,
+      "step": 639000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.147791295881192e-05,
+      "loss": 1.084,
+      "step": 640000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.1324584409263834e-05,
+      "loss": 1.0841,
+      "step": 641000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.117130650748039e-05,
+      "loss": 1.1045,
+      "step": 642000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.101838748185837e-05,
+      "loss": 1.1229,
+      "step": 643000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0865522860779585e-05,
+      "loss": 1.0369,
+      "step": 644000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.071286749127718e-05,
+      "loss": 0.8539,
+      "step": 645000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.056042304276582e-05,
+      "loss": 0.8873,
+      "step": 646000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.0408343307471943e-05,
+      "loss": 0.8589,
+      "step": 647000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.025632548485837e-05,
+      "loss": 0.8197,
+      "step": 648000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.010467526941344e-05,
+      "loss": 0.9677,
+      "step": 649000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9953090715801634e-05,
+      "loss": 0.9871,
+      "step": 650000
+    },
+    {
+      "epoch": 1.04,
+      "eval_accuracy": 0.831968594961707,
+      "eval_loss": 0.7390721440315247,
+      "eval_runtime": 10.7168,
+      "eval_samples_per_second": 466.556,
+      "eval_steps_per_second": 0.933,
+      "step": 650000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.980187664724534e-05,
+      "loss": 0.9091,
+      "step": 651000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.965073198680562e-05,
+      "loss": 0.9263,
+      "step": 652000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9499960673118322e-05,
+      "loss": 0.9273,
+      "step": 653000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.934926251079786e-05,
+      "loss": 0.8529,
+      "step": 654000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9198790181634074e-05,
+      "loss": 0.8239,
+      "step": 655000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9048695461845966e-05,
+      "loss": 0.7665,
+      "step": 656000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8898679503189996e-05,
+      "loss": 0.7491,
+      "step": 657000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8749043974581896e-05,
+      "loss": 0.8524,
+      "step": 658000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.859949094205726e-05,
+      "loss": 0.9649,
+      "step": 659000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.845017194207104e-05,
+      "loss": 0.9405,
+      "step": 660000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.83012375726286e-05,
+      "loss": 0.9965,
+      "step": 661000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.8152391295818637e-05,
+      "loss": 1.0081,
+      "step": 662000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.800393242842937e-05,
+      "loss": 0.981,
+      "step": 663000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7855565379304605e-05,
+      "loss": 0.8153,
+      "step": 664000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.7707588501547865e-05,
+      "loss": 0.8235,
+      "step": 665000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7559707163675296e-05,
+      "loss": 0.8797,
+      "step": 666000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7412071229249947e-05,
+      "loss": 0.8552,
+      "step": 667000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.7264829577788996e-05,
+      "loss": 0.978,
+      "step": 668000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.711768904169148e-05,
+      "loss": 1.0823,
+      "step": 669000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6970945507650934e-05,
+      "loss": 1.0495,
+      "step": 670000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6824306799850207e-05,
+      "loss": 1.0896,
+      "step": 671000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6678067795850213e-05,
+      "loss": 1.1019,
+      "step": 672000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.653193732438558e-05,
+      "loss": 1.0991,
+      "step": 673000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.6386209240998583e-05,
+      "loss": 0.9356,
+      "step": 674000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.624059339169965e-05,
+      "loss": 0.942,
+      "step": 675000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.6095237371464796e-05,
+      "loss": 0.9061,
+      "step": 676000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5950287733378777e-05,
+      "loss": 0.8134,
+      "step": 677000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.580545587338671e-05,
+      "loss": 0.9557,
+      "step": 678000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5661033035637218e-05,
+      "loss": 1.0279,
+      "step": 679000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.551673166496884e-05,
+      "loss": 1.0323,
+      "step": 680000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.537284193876279e-05,
+      "loss": 1.003,
+      "step": 681000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5229077363316923e-05,
+      "loss": 1.0201,
+      "step": 682000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.5085727036564444e-05,
+      "loss": 0.9572,
+      "step": 683000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.494250553878222e-05,
+      "loss": 0.7758,
+      "step": 684000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.4799558065744872e-05,
+      "loss": 0.7355,
+      "step": 685000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.465688618070303e-05,
+      "loss": 0.6992,
+      "step": 686000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.4514633699676327e-05,
+      "loss": 0.7698,
+      "step": 687000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.437251738882307e-05,
+      "loss": 0.9104,
+      "step": 688000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.4230823031561167e-05,
+      "loss": 0.8912,
+      "step": 689000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.408926850530927e-05,
+      "loss": 0.9452,
+      "step": 690000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.394813846378582e-05,
+      "loss": 1.0174,
+      "step": 691000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.380715190799833e-05,
+      "loss": 0.9929,
+      "step": 692000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3666592349526597e-05,
+      "loss": 0.8779,
+      "step": 693000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3526179925246578e-05,
+      "loss": 0.8655,
+      "step": 694000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.3386057014421453e-05,
+      "loss": 0.8112,
+      "step": 695000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3246364835391272e-05,
+      "loss": 0.7425,
+      "step": 696000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.3106825252040365e-05,
+      "loss": 0.7938,
+      "step": 697000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.296771886621397e-05,
+      "loss": 0.9602,
+      "step": 698000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2828768708148356e-05,
+      "loss": 1.0117,
+      "step": 699000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.269011569031495e-05,
+      "loss": 1.0679,
+      "step": 700000
+    },
+    {
+      "epoch": 1.09,
+      "eval_accuracy": 0.8310942232767082,
+      "eval_loss": 0.7435500025749207,
+      "eval_runtime": 12.3346,
+      "eval_samples_per_second": 405.365,
+      "eval_steps_per_second": 0.811,
+      "step": 700000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2551899533679438e-05,
+      "loss": 1.0913,
+      "step": 701000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2413845040981702e-05,
+      "loss": 1.0949,
+      "step": 702000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2276229827680116e-05,
+      "loss": 0.9394,
+      "step": 703000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.213891719102571e-05,
+      "loss": 0.9283,
+      "step": 704000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.2001771636780773e-05,
+      "loss": 0.8962,
+      "step": 705000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1864932266675498e-05,
+      "loss": 0.8581,
+      "step": 706000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1728400577163317e-05,
+      "loss": 0.9983,
+      "step": 707000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1592314128920388e-05,
+      "loss": 1.0883,
+      "step": 708000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1456401965073002e-05,
+      "loss": 1.0622,
+      "step": 709000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1320937393044398e-05,
+      "loss": 1.0325,
+      "step": 710000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.118565069415642e-05,
+      "loss": 1.0107,
+      "step": 711000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.1050813918077206e-05,
+      "loss": 1.0555,
+      "step": 712000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.0916158596110753e-05,
+      "loss": 0.8862,
+      "step": 713000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.0781821330374222e-05,
+      "loss": 0.8341,
+      "step": 714000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.0647937447607897e-05,
+      "loss": 0.8047,
+      "step": 715000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.0514240376387622e-05,
+      "loss": 0.7639,
+      "step": 716000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.0380998969779935e-05,
+      "loss": 0.8598,
+      "step": 717000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.0247947935572087e-05,
+      "loss": 0.8621,
+      "step": 718000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.0115354827706882e-05,
+      "loss": 0.8859,
+      "step": 719000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9982955644776074e-05,
+      "loss": 0.9538,
+      "step": 720000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9851016629906853e-05,
+      "loss": 0.9508,
+      "step": 721000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9719275084032178e-05,
+      "loss": 0.8639,
+      "step": 722000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.958786468331379e-05,
+      "loss": 0.8306,
+      "step": 723000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.9456917776051037e-05,
+      "loss": 0.8304,
+      "step": 724000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.932617363852691e-05,
+      "loss": 0.7599,
+      "step": 725000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.9195895185717727e-05,
+      "loss": 0.8071,
+      "step": 726000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.906582302484229e-05,
+      "loss": 0.9583,
+      "step": 727000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8936089155170195e-05,
+      "loss": 0.9304,
+      "step": 728000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8806824219453507e-05,
+      "loss": 0.9371,
+      "step": 729000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8677770842888893e-05,
+      "loss": 0.9491,
+      "step": 730000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8549188540480782e-05,
+      "loss": 1.0003,
+      "step": 731000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8420821296592628e-05,
+      "loss": 0.9365,
+      "step": 732000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8292799397543353e-05,
+      "loss": 0.9281,
+      "step": 733000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8165251744849992e-05,
+      "loss": 0.8999,
+      "step": 734000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8037924382930186e-05,
+      "loss": 0.8513,
+      "step": 735000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7911073355924523e-05,
+      "loss": 0.9405,
+      "step": 736000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7784446095277167e-05,
+      "loss": 1.0675,
+      "step": 737000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7658297237286585e-05,
+      "loss": 1.0324,
+      "step": 738000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7532375611460606e-05,
+      "loss": 1.0777,
+      "step": 739000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.740693443512806e-05,
+      "loss": 1.1039,
+      "step": 740000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7281723946836492e-05,
+      "loss": 1.1023,
+      "step": 741000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.715699593387959e-05,
+      "loss": 0.9438,
+      "step": 742000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.7032502054759063e-05,
+      "loss": 0.8537,
+      "step": 743000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.6908492655730533e-05,
+      "loss": 0.8055,
+      "step": 744000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.6784720826102262e-05,
+      "loss": 0.8227,
+      "step": 745000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.666143546015176e-05,
+      "loss": 0.9655,
+      "step": 746000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.6538391088783723e-05,
+      "loss": 0.9601,
+      "step": 747000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.6415835143421003e-05,
+      "loss": 0.9641,
+      "step": 748000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.629352360729162e-05,
+      "loss": 0.9631,
+      "step": 749000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.6171580679787825e-05,
+      "loss": 0.9203,
+      "step": 750000
+    },
+    {
+      "epoch": 1.14,
+      "eval_accuracy": 0.8355373765019023,
+      "eval_loss": 0.7186969518661499,
+      "eval_runtime": 10.9844,
+      "eval_samples_per_second": 455.192,
+      "eval_steps_per_second": 0.91,
+      "step": 750000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.6050007694457925e-05,
+      "loss": 0.8906,
+      "step": 751000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.592892699662655e-05,
+      "loss": 0.7487,
+      "step": 752000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5808097506834524e-05,
+      "loss": 0.7857,
+      "step": 753000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.5687641934210118e-05,
+      "loss": 0.7229,
+      "step": 754000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.556768148850735e-05,
+      "loss": 0.7239,
+      "step": 755000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.544797732076107e-05,
+      "loss": 0.923,
+      "step": 756000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.532877014553799e-05,
+      "loss": 0.9589,
+      "step": 757000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.5209822613682983e-05,
+      "loss": 0.9368,
+      "step": 758000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.5091255541633964e-05,
+      "loss": 0.9942,
+      "step": 759000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4973188220219254e-05,
+      "loss": 0.996,
+      "step": 760000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4855385569805891e-05,
+      "loss": 0.898,
+      "step": 761000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4738084481176312e-05,
+      "loss": 0.8023,
+      "step": 762000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4621051400316382e-05,
+      "loss": 0.7635,
+      "step": 763000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4504405217970129e-05,
+      "loss": 0.7792,
+      "step": 764000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4388263273453235e-05,
+      "loss": 0.8774,
+      "step": 765000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4272509994685329e-05,
+      "loss": 1.0395,
+      "step": 766000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4157031361942913e-05,
+      "loss": 1.0266,
+      "step": 767000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4041944702162985e-05,
+      "loss": 1.0627,
+      "step": 768000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3927251273914792e-05,
+      "loss": 1.083,
+      "step": 769000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3813066432947708e-05,
+      "loss": 1.0811,
+      "step": 770000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3699162829897188e-05,
+      "loss": 0.9505,
+      "step": 771000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.358576951490385e-05,
+      "loss": 0.9315,
+      "step": 772000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3472660714582335e-05,
+      "loss": 0.9083,
+      "step": 773000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3360063880794788e-05,
+      "loss": 0.8656,
+      "step": 774000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3247754826001119e-05,
+      "loss": 0.9627,
+      "step": 775000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3135847687872443e-05,
+      "loss": 0.9883,
+      "step": 776000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3024454992430079e-05,
+      "loss": 0.9516,
+      "step": 777000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2913354949650841e-05,
+      "loss": 1.044,
+      "step": 778000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2802881463850613e-05,
+      "loss": 1.0429,
+      "step": 779000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2692592964051836e-05,
+      "loss": 0.9851,
+      "step": 780000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2582712452079226e-05,
+      "loss": 0.8493,
+      "step": 781000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2473241129568458e-05,
+      "loss": 0.7932,
+      "step": 782000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.236428904923082e-05,
+      "loss": 0.702,
+      "step": 783000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2255639280464832e-05,
+      "loss": 0.6618,
+      "step": 784000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.2147402277980474e-05,
+      "loss": 0.8469,
+      "step": 785000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.2039686841331998e-05,
+      "loss": 0.9126,
+      "step": 786000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1932278502155054e-05,
+      "loss": 0.8954,
+      "step": 787000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1825286465481434e-05,
+      "loss": 0.9188,
+      "step": 788000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1718818267007175e-05,
+      "loss": 0.9873,
+      "step": 789000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1612661921699398e-05,
+      "loss": 0.9544,
+      "step": 790000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.150703090064395e-05,
+      "loss": 0.801,
+      "step": 791000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1401714885682025e-05,
+      "loss": 0.8218,
+      "step": 792000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1296820975382121e-05,
+      "loss": 0.7743,
+      "step": 793000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1192454575710875e-05,
+      "loss": 0.7675,
+      "step": 794000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.1088407886452029e-05,
+      "loss": 0.9008,
+      "step": 795000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0984890136358416e-05,
+      "loss": 0.8879,
+      "step": 796000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0881695214929688e-05,
+      "loss": 0.9673,
+      "step": 797000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0778928085014794e-05,
+      "loss": 1.0564,
+      "step": 798000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0676691994057019e-05,
+      "loss": 1.0726,
+      "step": 799000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0574783383421865e-05,
+      "loss": 0.9626,
+      "step": 800000
+    },
+    {
+      "epoch": 1.19,
+      "eval_accuracy": 0.8352768932764173,
+      "eval_loss": 0.7242327928543091,
+      "eval_runtime": 12.0402,
+      "eval_samples_per_second": 415.274,
+      "eval_steps_per_second": 0.831,
+      "step": 800000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0473407182373813e-05,
+      "loss": 0.9342,
+      "step": 801000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0372361544374464e-05,
+      "loss": 0.8919,
+      "step": 802000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0271849663326171e-05,
+      "loss": 0.8567,
+      "step": 803000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0171671413607247e-05,
+      "loss": 0.9096,
+      "step": 804000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.0072028244878407e-05,
+      "loss": 1.0644,
+      "step": 805000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.97272176118008e-06,
+      "loss": 1.0327,
+      "step": 806000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.873853009383633e-06,
+      "loss": 1.076,
+      "step": 807000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.775423070701888e-06,
+      "loss": 1.0958,
+      "step": 808000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.67753079151617e-06,
+      "loss": 1.0053,
+      "step": 809000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.57998126200062e-06,
+      "loss": 0.8648,
+      "step": 810000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.482970645690526e-06,
+      "loss": 0.8919,
+      "step": 811000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.38630578820755e-06,
+      "loss": 0.8439,
+      "step": 812000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.290181074089233e-06,
+      "loss": 0.7462,
+      "step": 813000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.194405112845483e-06,
+      "loss": 0.8975,
+      "step": 814000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.099170501521598e-06,
+      "loss": 0.9167,
+      "step": 815000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.00428762187942e-06,
+      "loss": 0.8518,
+      "step": 816000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.909947275055568e-06,
+      "loss": 0.8843,
+      "step": 817000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.815961623350038e-06,
+      "loss": 0.926,
+      "step": 818000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.722519663652901e-06,
+      "loss": 0.9408,
+      "step": 819000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.629435347010716e-06,
+      "loss": 0.7628,
+      "step": 820000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.536803452235437e-06,
+      "loss": 0.7669,
+      "step": 821000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.444624992334588e-06,
+      "loss": 0.7749,
+      "step": 822000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.352992472045557e-06,
+      "loss": 0.7575,
+      "step": 823000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.26172344512513e-06,
+      "loss": 0.899,
+      "step": 824000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 8.171001445569593e-06,
+      "loss": 0.9519,
+      "step": 825000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 8.080645840041112e-06,
+      "loss": 0.9472,
+      "step": 826000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.990838325725758e-06,
+      "loss": 0.9576,
+      "step": 827000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.901400090084665e-06,
+      "loss": 0.9493,
+      "step": 828000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.81251098555364e-06,
+      "loss": 0.8743,
+      "step": 829000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.723994028206778e-06,
+      "loss": 0.8593,
+      "step": 830000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.636027217870157e-06,
+      "loss": 0.8767,
+      "step": 831000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.54843540696496e-06,
+      "loss": 0.853,
+      "step": 832000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.461394734929022e-06,
+      "loss": 0.8742,
+      "step": 833000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.374731898184495e-06,
+      "loss": 1.0273,
+      "step": 834000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 7.2886211680837424e-06,
+      "loss": 1.0271,
+      "step": 835000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.202891092623126e-06,
+      "loss": 1.049,
+      "step": 836000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.11771406745404e-06,
+      "loss": 1.0773,
+      "step": 837000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 7.032920499639423e-06,
+      "loss": 1.0882,
+      "step": 838000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.94868090159605e-06,
+      "loss": 0.9609,
+      "step": 839000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.864827546864583e-06,
+      "loss": 0.9468,
+      "step": 840000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.781529057175845e-06,
+      "loss": 0.8922,
+      "step": 841000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.698619579877818e-06,
+      "loss": 0.7689,
+      "step": 842000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.616183639538559e-06,
+      "loss": 0.852,
+      "step": 843000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.534385586581854e-06,
+      "loss": 1.0235,
+      "step": 844000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 6.452898467929852e-06,
+      "loss": 0.9675,
+      "step": 845000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.371887573403335e-06,
+      "loss": 0.9587,
+      "step": 846000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.29135378892447e-06,
+      "loss": 1.0189,
+      "step": 847000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.211377811943364e-06,
+      "loss": 0.9343,
+      "step": 848000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.131879743067948e-06,
+      "loss": 0.7473,
+      "step": 849000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 6.0528604499385185e-06,
+      "loss": 0.7263,
+      "step": 850000
+    },
+    {
+      "epoch": 1.24,
+      "eval_accuracy": 0.8377542903972233,
+      "eval_loss": 0.7093546986579895,
+      "eval_runtime": 11.2849,
+      "eval_samples_per_second": 443.071,
+      "eval_steps_per_second": 0.886,
+      "step": 850000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.974242417283055e-06,
+      "loss": 0.7407,
+      "step": 851000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.896105843258959e-06,
+      "loss": 0.7019,
+      "step": 852000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.818451582355289e-06,
+      "loss": 0.8287,
+      "step": 853000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.741280483786604e-06,
+      "loss": 0.9027,
+      "step": 854000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.664593391483675e-06,
+      "loss": 0.9336,
+      "step": 855000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.588467103873829e-06,
+      "loss": 0.9669,
+      "step": 856000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5127500486207735e-06,
+      "loss": 0.9644,
+      "step": 857000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.437594486071778e-06,
+      "loss": 0.9871,
+      "step": 858000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.362850776666601e-06,
+      "loss": 0.7998,
+      "step": 859000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.288595211973824e-06,
+      "loss": 0.799,
+      "step": 860000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.214902126145504e-06,
+      "loss": 0.7326,
+      "step": 861000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.141624791506633e-06,
+      "loss": 0.7172,
+      "step": 862000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.068910562340645e-06,
+      "loss": 0.956,
+      "step": 863000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.9966146599391025e-06,
+      "loss": 1.0129,
+      "step": 864000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.924882465011432e-06,
+      "loss": 1.0269,
+      "step": 865000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.853571154142544e-06,
+      "loss": 1.0608,
+      "step": 866000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.782824128114777e-06,
+      "loss": 1.0748,
+      "step": 867000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.7125005250476976e-06,
+      "loss": 0.9914,
+      "step": 868000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.642741759530028e-06,
+      "loss": 0.9261,
+      "step": 869000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.573408937371581e-06,
+      "loss": 0.8934,
+      "step": 870000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.504572893135017e-06,
+      "loss": 0.868,
+      "step": 871000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.436302469348041e-06,
+      "loss": 0.8991,
+      "step": 872000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.368461735204643e-06,
+      "loss": 1.0529,
+      "step": 873000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.3011871124621965e-06,
+      "loss": 1.0182,
+      "step": 874000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.23434465301864e-06,
+      "loss": 0.9739,
+      "step": 875000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.168068771188594e-06,
+      "loss": 1.0152,
+      "step": 876000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.10222750742742e-06,
+      "loss": 1.0456,
+      "step": 877000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.036888174352743e-06,
+      "loss": 0.8894,
+      "step": 878000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.972116071885162e-06,
+      "loss": 0.8262,
+      "step": 879000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.907782234601492e-06,
+      "loss": 0.8263,
+      "step": 880000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.844016032197534e-06,
+      "loss": 0.6921,
+      "step": 881000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.78069050206184e-06,
+      "loss": 0.75,
+      "step": 882000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.7179329862731317e-06,
+      "loss": 0.8619,
+      "step": 883000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.6556185305799074e-06,
+      "loss": 0.8767,
+      "step": 884000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.593872443878982e-06,
+      "loss": 0.92,
+      "step": 885000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.532571785739236e-06,
+      "loss": 0.9303,
+      "step": 886000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.4718398263996642e-06,
+      "loss": 0.9353,
+      "step": 887000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.411555644621961e-06,
+      "loss": 0.834,
+      "step": 888000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.3517809466076076e-06,
+      "loss": 0.8077,
+      "step": 889000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.2925753955738435e-06,
+      "loss": 0.7661,
+      "step": 890000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.2338211094595227e-06,
+      "loss": 0.7519,
+      "step": 891000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.1755782507849554e-06,
+      "loss": 0.8897,
+      "step": 892000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.117904931292237e-06,
+      "loss": 0.9311,
+      "step": 893000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.060686319691808e-06,
+      "loss": 0.8915,
+      "step": 894000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.004037477585625e-06,
+      "loss": 0.9306,
+      "step": 895000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.9478456134844844e-06,
+      "loss": 1.0244,
+      "step": 896000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.892223724291321e-06,
+      "loss": 1.029,
+      "step": 897000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8370610632462602e-06,
+      "loss": 0.8896,
+      "step": 898000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.7824685576152752e-06,
+      "loss": 0.8942,
+      "step": 899000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.728337510207396e-06,
+      "loss": 0.8578,
+      "step": 900000
+    },
+    {
+      "epoch": 1.29,
+      "eval_accuracy": 0.8368069978989446,
+      "eval_loss": 0.7140042185783386,
+      "eval_runtime": 10.5028,
+      "eval_samples_per_second": 476.062,
+      "eval_steps_per_second": 0.952,
+      "step": 900000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.6747767738041995e-06,
+      "loss": 0.861,
+      "step": 901000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.621679705533614e-06,
+      "loss": 1.0138,
+      "step": 902000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.569100758983095e-06,
+      "loss": 1.0312,
+      "step": 903000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.517092310118435e-06,
+      "loss": 1.0481,
+      "step": 904000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.4655508067719378e-06,
+      "loss": 1.0777,
+      "step": 905000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.4145798943795206e-06,
+      "loss": 1.0943,
+      "step": 906000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.3640780865753555e-06,
+      "loss": 0.9609,
+      "step": 907000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.314146938058942e-06,
+      "loss": 0.8363,
+      "step": 908000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.2646870327012294e-06,
+      "loss": 0.842,
+      "step": 909000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.215797830027422e-06,
+      "loss": 0.8333,
+      "step": 910000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.1673819884896694e-06,
+      "loss": 0.8255,
+      "step": 911000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.119489236799016e-06,
+      "loss": 0.9393,
+      "step": 912000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.0721200987029453e-06,
+      "loss": 0.9491,
+      "step": 913000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.025321675254488e-06,
+      "loss": 0.8986,
+      "step": 914000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.9790007877817653e-06,
+      "loss": 0.8977,
+      "step": 915000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.933250583520735e-06,
+      "loss": 0.8944,
+      "step": 916000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.887979970871323e-06,
+      "loss": 0.8071,
+      "step": 917000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.8432355035647219e-06,
+      "loss": 0.7774,
+      "step": 918000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.799061625536469e-06,
+      "loss": 0.7421,
+      "step": 919000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.7553703837534352e-06,
+      "loss": 0.7138,
+      "step": 920000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.712249637465102e-06,
+      "loss": 0.86,
+      "step": 921000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.6696135304913818e-06,
+      "loss": 0.9401,
+      "step": 922000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.627547800298329e-06,
+      "loss": 0.9108,
+      "step": 923000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.5859686913107963e-06,
+      "loss": 0.9746,
+      "step": 924000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.5449190304130446e-06,
+      "loss": 0.975,
+      "step": 925000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.5044395214499573e-06,
+      "loss": 0.9505,
+      "step": 926000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4644495671149062e-06,
+      "loss": 0.7784,
+      "step": 927000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.4250295836934214e-06,
+      "loss": 0.7832,
+      "step": 928000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3861010834623766e-06,
+      "loss": 0.8266,
+      "step": 929000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3477423482163465e-06,
+      "loss": 0.8174,
+      "step": 930000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3098770032456076e-06,
+      "loss": 0.9858,
+      "step": 931000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.272581192432859e-06,
+      "loss": 1.0364,
+      "step": 932000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2357806574193386e-06,
+      "loss": 1.029,
+      "step": 933000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.1995494008484487e-06,
+      "loss": 1.0567,
+      "step": 934000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.1638152839573868e-06,
+      "loss": 1.0777,
+      "step": 935000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.12861523175814e-06,
+      "loss": 1.0355,
+      "step": 936000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.0939840277111369e-06,
+      "loss": 0.9177,
+      "step": 937000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.0598527188610874e-06,
+      "loss": 0.9093,
+      "step": 938000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.0262899402720737e-06,
+      "loss": 0.8764,
+      "step": 939000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.93228866292617e-07,
+      "loss": 0.8315,
+      "step": 940000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.607359798384785e-07,
+      "loss": 0.943,
+      "step": 941000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.287465854822597e-07,
+      "loss": 0.9774,
+      "step": 942000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.973250110861309e-07,
+      "loss": 1.0164,
+      "step": 943000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.664086942739291e-07,
+      "loss": 0.9901,
+      "step": 944000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 8.360296945443436e-07,
+      "loss": 0.9853,
+      "step": 945000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 8.061883441172646e-07,
+      "loss": 0.8922,
+      "step": 946000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 7.769140038817014e-07,
+      "loss": 0.7675,
+      "step": 947000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 7.48148386742864e-07,
+      "loss": 0.7123,
+      "step": 948000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 7.199213799624294e-07,
+      "loss": 0.6682,
+      "step": 949000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.922607110228129e-07,
+      "loss": 0.7693,
+      "step": 950000
+    },
+    {
+      "epoch": 1.34,
+      "eval_accuracy": 0.8377192585963231,
+      "eval_loss": 0.7091149091720581,
+      "eval_runtime": 9.9645,
+      "eval_samples_per_second": 501.78,
+      "eval_steps_per_second": 1.004,
+      "step": 950000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.651113057525916e-07,
+      "loss": 0.899,
+      "step": 951000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.385277592210082e-07,
+      "loss": 0.8642,
+      "step": 952000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 6.124571418766378e-07,
+      "loss": 0.9313,
+      "step": 953000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.869518794409723e-07,
+      "loss": 0.9976,
+      "step": 954000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.619611892955956e-07,
+      "loss": 0.9632,
+      "step": 955000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.375353255232474e-07,
+      "loss": 0.833,
+      "step": 956000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.136256546577067e-07,
+      "loss": 0.8247,
+      "step": 957000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.902571588535909e-07,
+      "loss": 0.789,
+      "step": 958000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.674526502020382e-07,
+      "loss": 0.7219,
+      "step": 959000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.451667234591728e-07,
+      "loss": 0.8131,
+      "step": 960000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.234441936661282e-07,
+      "loss": 0.9334,
+      "step": 961000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.0224180986853655e-07,
+      "loss": 0.9777,
+      "step": 962000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.8160220819785095e-07,
+      "loss": 1.0358,
+      "step": 963000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.6148429387927175e-07,
+      "loss": 1.0623,
+      "step": 964000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.419285222713675e-07,
+      "loss": 1.0595,
+      "step": 965000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.228959565747369e-07,
+      "loss": 0.9013,
+      "step": 966000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.044248696072116e-07,
+      "loss": 0.9035,
+      "step": 967000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.86478484246272e-07,
+      "loss": 0.865,
+      "step": 968000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.690928890965172e-07,
+      "loss": 0.8449,
+      "step": 969000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.5223346831947934e-07,
+      "loss": 1.0131,
+      "step": 970000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.359180811469297e-07,
+      "loss": 1.0522,
+      "step": 971000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.2016240528467956e-07,
+      "loss": 1.0527,
+      "step": 972000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.0493507016841605e-07,
+      "loss": 1.0244,
+      "step": 973000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.902666966315303e-07,
+      "loss": 0.9939,
+      "step": 974000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.7612807899859974e-07,
+      "loss": 0.975,
+      "step": 975000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.6254764878778085e-07,
+      "loss": 0.8924,
+      "step": 976000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.4949836648880388e-07,
+      "loss": 0.8095,
+      "step": 977000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3699424128894024e-07,
+      "loss": 0.7741,
+      "step": 978000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.2503540993129005e-07,
+      "loss": 0.7888,
+      "step": 979000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1363314412082271e-07,
+      "loss": 0.8592,
+      "step": 980000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.0276474121272417e-07,
+      "loss": 0.8308,
+      "step": 981000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.245205661059241e-08,
+      "loss": 0.8657,
+      "step": 982000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 8.2674557095902e-08,
+      "loss": 0.9413,
+      "step": 983000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 7.345190436134352e-08,
+      "loss": 0.9376,
+      "step": 984000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 6.476573551197352e-08,
+      "loss": 0.8103,
+      "step": 985000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.6633517670373616e-08,
+      "loss": 0.8122,
+      "step": 986000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.903905906762374e-08,
+      "loss": 0.809,
+      "step": 987000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.1990856170864845e-08,
+      "loss": 0.7404,
+      "step": 988000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.549521501085562e-08,
+      "loss": 0.8188,
+      "step": 989000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.9539202348127794e-08,
+      "loss": 0.9535,
+      "step": 990000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.412965863871075e-08,
+      "loss": 0.9178,
+      "step": 991000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.9271233047113424e-08,
+      "loss": 0.9305,
+      "step": 992000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.4954252135407352e-08,
+      "loss": 0.938,
+      "step": 993000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1187396973016962e-08,
+      "loss": 0.969,
+      "step": 994000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 7.963167508967528e-09,
+      "loss": 0.8864,
+      "step": 995000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.288047385498818e-09,
+      "loss": 0.8899,
+      "step": 996000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.156710266344343e-09,
+      "loss": 0.8701,
+      "step": 997000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.5721306152016724e-09,
+      "loss": 0.8305,
+      "step": 998000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.350904473455653e-10,
+      "loss": 0.961,
+      "step": 999000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.3524949094875254e-11,
+      "loss": 1.0488,
+      "step": 1000000
+    },
+    {
+      "epoch": 1.39,
+      "eval_accuracy": 0.8387476162115661,
+      "eval_loss": 0.7079769372940063,
+      "eval_runtime": 10.2424,
+      "eval_samples_per_second": 488.169,
+      "eval_steps_per_second": 0.976,
+      "step": 1000000
+    },
+    {
+      "epoch": 1.39,
+      "step": 1000000,
+      "total_flos": 6.738157336654774e+19,
+      "train_loss": 0.18125205249023438,
+      "train_runtime": 269324.8897,
+      "train_samples_per_second": 3802.099,
+      "train_steps_per_second": 3.713
+    }
+  ],
+  "max_steps": 1000000,
+  "num_train_epochs": 9223372036854775807,
+  "total_flos": 6.738157336654774e+19,
+  "trial_name": null,
+  "trial_params": null
+}