diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14856 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.9736303748052775,
+  "global_step": 530000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9953078958728256e-05,
+      "loss": 6.7877,
+      "step": 500
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 6.12216329574585,
+      "eval_runtime": 227.3514,
+      "eval_samples_per_second": 439.848,
+      "eval_steps_per_second": 13.745,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.990615791745651e-05,
+      "loss": 5.9665,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 5.832594394683838,
+      "eval_runtime": 227.8781,
+      "eval_samples_per_second": 438.831,
+      "eval_steps_per_second": 13.713,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9859236876184755e-05,
+      "loss": 5.7813,
+      "step": 1500
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 5.70578145980835,
+      "eval_runtime": 227.8691,
+      "eval_samples_per_second": 438.848,
+      "eval_steps_per_second": 13.714,
+      "step": 1500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.981231583491301e-05,
+      "loss": 5.6811,
+      "step": 2000
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 5.598133563995361,
+      "eval_runtime": 228.0113,
+      "eval_samples_per_second": 438.575,
+      "eval_steps_per_second": 13.705,
+      "step": 2000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.976539479364126e-05,
+      "loss": 5.5624,
+      "step": 2500
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 5.457843780517578,
+      "eval_runtime": 230.6878,
+      "eval_samples_per_second": 433.486,
+      "eval_steps_per_second": 13.546,
+      "step": 2500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9718473752369515e-05,
+      "loss": 5.4101,
+      "step": 3000
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 5.203133583068848,
+      "eval_runtime": 230.8416,
+      "eval_samples_per_second": 433.198,
+      "eval_steps_per_second": 13.537,
+      "step": 3000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.967155271109777e-05,
+      "loss": 5.1461,
+      "step": 3500
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.861812591552734,
+      "eval_runtime": 231.3953,
+      "eval_samples_per_second": 432.161,
+      "eval_steps_per_second": 13.505,
+      "step": 3500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.962463166982602e-05,
+      "loss": 4.8415,
+      "step": 4000
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 4.487979888916016,
+      "eval_runtime": 231.5686,
+      "eval_samples_per_second": 431.837,
+      "eval_steps_per_second": 13.495,
+      "step": 4000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.957771062855427e-05,
+      "loss": 4.5424,
+      "step": 4500
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 4.209322452545166,
+      "eval_runtime": 232.2861,
+      "eval_samples_per_second": 430.504,
+      "eval_steps_per_second": 13.453,
+      "step": 4500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.953078958728252e-05,
+      "loss": 4.286,
+      "step": 5000
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 3.9922587871551514,
+      "eval_runtime": 231.4165,
+      "eval_samples_per_second": 432.121,
+      "eval_steps_per_second": 13.504,
+      "step": 5000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9483868546010775e-05,
+      "loss": 4.0996,
+      "step": 5500
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 3.8145012855529785,
+      "eval_runtime": 230.7704,
+      "eval_samples_per_second": 433.331,
+      "eval_steps_per_second": 13.542,
+      "step": 5500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.943694750473903e-05,
+      "loss": 3.9396,
+      "step": 6000
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 3.652764320373535,
+      "eval_runtime": 230.0751,
+      "eval_samples_per_second": 434.641,
+      "eval_steps_per_second": 13.583,
+      "step": 6000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.939002646346728e-05,
+      "loss": 3.8204,
+      "step": 6500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 3.5459558963775635,
+      "eval_runtime": 227.8247,
+      "eval_samples_per_second": 438.934,
+      "eval_steps_per_second": 13.717,
+      "step": 6500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.934310542219553e-05,
+      "loss": 3.7161,
+      "step": 7000
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 3.432518720626831,
+      "eval_runtime": 228.0814,
+      "eval_samples_per_second": 438.44,
+      "eval_steps_per_second": 13.701,
+      "step": 7000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.929618438092378e-05,
+      "loss": 3.602,
+      "step": 7500
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 3.3445019721984863,
+      "eval_runtime": 227.7955,
+      "eval_samples_per_second": 438.99,
+      "eval_steps_per_second": 13.718,
+      "step": 7500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9249263339652035e-05,
+      "loss": 3.519,
+      "step": 8000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 3.2647743225097656,
+      "eval_runtime": 227.9434,
+      "eval_samples_per_second": 438.705,
+      "eval_steps_per_second": 13.71,
+      "step": 8000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.920234229838029e-05,
+      "loss": 3.4347,
+      "step": 8500
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 3.188307762145996,
+      "eval_runtime": 228.5169,
+      "eval_samples_per_second": 437.604,
+      "eval_steps_per_second": 13.675,
+      "step": 8500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.915542125710854e-05,
+      "loss": 3.3642,
+      "step": 9000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 3.1089322566986084,
+      "eval_runtime": 228.7192,
+      "eval_samples_per_second": 437.217,
+      "eval_steps_per_second": 13.663,
+      "step": 9000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9108500215836794e-05,
+      "loss": 3.29,
+      "step": 9500
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 3.044532299041748,
+      "eval_runtime": 229.7935,
+      "eval_samples_per_second": 435.173,
+      "eval_steps_per_second": 13.599,
+      "step": 9500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.906157917456504e-05,
+      "loss": 3.2545,
+      "step": 10000
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 3.0254080295562744,
+      "eval_runtime": 229.6489,
+      "eval_samples_per_second": 435.447,
+      "eval_steps_per_second": 13.608,
+      "step": 10000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9014658133293294e-05,
+      "loss": 3.2036,
+      "step": 10500
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.9655680656433105,
+      "eval_runtime": 228.7959,
+      "eval_samples_per_second": 437.071,
+      "eval_steps_per_second": 13.658,
+      "step": 10500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.896773709202155e-05,
+      "loss": 3.1292,
+      "step": 11000
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.92611026763916,
+      "eval_runtime": 224.7807,
+      "eval_samples_per_second": 444.878,
+      "eval_steps_per_second": 13.902,
+      "step": 11000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.89208160507498e-05,
+      "loss": 3.1273,
+      "step": 11500
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.880321502685547,
+      "eval_runtime": 225.1803,
+      "eval_samples_per_second": 444.089,
+      "eval_steps_per_second": 13.878,
+      "step": 11500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8873895009478054e-05,
+      "loss": 3.0435,
+      "step": 12000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.8387928009033203,
+      "eval_runtime": 229.766,
+      "eval_samples_per_second": 435.225,
+      "eval_steps_per_second": 13.601,
+      "step": 12000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.88269739682063e-05,
+      "loss": 2.9999,
+      "step": 12500
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.7962071895599365,
+      "eval_runtime": 226.0122,
+      "eval_samples_per_second": 442.454,
+      "eval_steps_per_second": 13.827,
+      "step": 12500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8780052926934554e-05,
+      "loss": 2.9642,
+      "step": 13000
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 2.7659239768981934,
+      "eval_runtime": 225.1957,
+      "eval_samples_per_second": 444.058,
+      "eval_steps_per_second": 13.877,
+      "step": 13000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.873313188566281e-05,
+      "loss": 2.9657,
+      "step": 13500
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.729302167892456,
+      "eval_runtime": 225.23,
+      "eval_samples_per_second": 443.99,
+      "eval_steps_per_second": 13.875,
+      "step": 13500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.868621084439106e-05,
+      "loss": 2.9025,
+      "step": 14000
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 2.698721170425415,
+      "eval_runtime": 225.2627,
+      "eval_samples_per_second": 443.926,
+      "eval_steps_per_second": 13.873,
+      "step": 14000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.8639289803119314e-05,
+      "loss": 2.9036,
+      "step": 14500
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.668630361557007,
+      "eval_runtime": 225.2216,
+      "eval_samples_per_second": 444.007,
+      "eval_steps_per_second": 13.875,
+      "step": 14500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.859236876184757e-05,
+      "loss": 2.8405,
+      "step": 15000
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 2.6363720893859863,
+      "eval_runtime": 226.5807,
+      "eval_samples_per_second": 441.344,
+      "eval_steps_per_second": 13.792,
+      "step": 15000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.8545447720575814e-05,
+      "loss": 2.8067,
+      "step": 15500
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.6151633262634277,
+      "eval_runtime": 230.6298,
+      "eval_samples_per_second": 433.595,
+      "eval_steps_per_second": 13.55,
+      "step": 15500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.849852667930407e-05,
+      "loss": 2.7782,
+      "step": 16000
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.586013078689575,
+      "eval_runtime": 237.3092,
+      "eval_samples_per_second": 421.391,
+      "eval_steps_per_second": 13.168,
+      "step": 16000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.845160563803232e-05,
+      "loss": 2.7647,
+      "step": 16500
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 2.5545260906219482,
+      "eval_runtime": 239.3104,
+      "eval_samples_per_second": 417.867,
+      "eval_steps_per_second": 13.058,
+      "step": 16500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.8404684596760573e-05,
+      "loss": 2.7115,
+      "step": 17000
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.5350353717803955,
+      "eval_runtime": 242.6234,
+      "eval_samples_per_second": 412.161,
+      "eval_steps_per_second": 12.88,
+      "step": 17000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.835776355548883e-05,
+      "loss": 2.6993,
+      "step": 17500
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 2.502089738845825,
+      "eval_runtime": 236.8183,
+      "eval_samples_per_second": 422.265,
+      "eval_steps_per_second": 13.196,
+      "step": 17500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.831084251421707e-05,
+      "loss": 2.7097,
+      "step": 18000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.493856191635132,
+      "eval_runtime": 236.3441,
+      "eval_samples_per_second": 423.112,
+      "eval_steps_per_second": 13.222,
+      "step": 18000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.8263921472945327e-05,
+      "loss": 2.6829,
+      "step": 18500
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.484680652618408,
+      "eval_runtime": 234.9827,
+      "eval_samples_per_second": 425.563,
+      "eval_steps_per_second": 13.299,
+      "step": 18500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.821700043167358e-05,
+      "loss": 2.6755,
+      "step": 19000
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.461320161819458,
+      "eval_runtime": 233.7755,
+      "eval_samples_per_second": 427.761,
+      "eval_steps_per_second": 13.368,
+      "step": 19000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.817007939040183e-05,
+      "loss": 2.6365,
+      "step": 19500
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 2.435065269470215,
+      "eval_runtime": 238.7041,
+      "eval_samples_per_second": 418.929,
+      "eval_steps_per_second": 13.092,
+      "step": 19500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8123158349130086e-05,
+      "loss": 2.6216,
+      "step": 20000
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.4164962768554688,
+      "eval_runtime": 234.7619,
+      "eval_samples_per_second": 425.964,
+      "eval_steps_per_second": 13.311,
+      "step": 20000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.807623730785834e-05,
+      "loss": 2.6014,
+      "step": 20500
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.402653932571411,
+      "eval_runtime": 234.6149,
+      "eval_samples_per_second": 426.23,
+      "eval_steps_per_second": 13.32,
+      "step": 20500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8029316266586586e-05,
+      "loss": 2.5833,
+      "step": 21000
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.3744804859161377,
+      "eval_runtime": 231.112,
+      "eval_samples_per_second": 432.691,
+      "eval_steps_per_second": 13.522,
+      "step": 21000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.798239522531484e-05,
+      "loss": 2.5663,
+      "step": 21500
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 2.35831880569458,
+      "eval_runtime": 233.0868,
+      "eval_samples_per_second": 429.025,
+      "eval_steps_per_second": 13.407,
+      "step": 21500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.793547418404309e-05,
+      "loss": 2.5626,
+      "step": 22000
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.3559041023254395,
+      "eval_runtime": 234.7406,
+      "eval_samples_per_second": 426.002,
+      "eval_steps_per_second": 13.313,
+      "step": 22000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.7888553142771346e-05,
+      "loss": 2.5453,
+      "step": 22500
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 2.3476722240448,
+      "eval_runtime": 231.9484,
+      "eval_samples_per_second": 431.13,
+      "eval_steps_per_second": 13.473,
+      "step": 22500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.78416321014996e-05,
+      "loss": 2.492,
+      "step": 23000
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.3358097076416016,
+      "eval_runtime": 231.7601,
+      "eval_samples_per_second": 431.481,
+      "eval_steps_per_second": 13.484,
+      "step": 23000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.779471106022785e-05,
+      "loss": 2.4961,
+      "step": 23500
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.31193208694458,
+      "eval_runtime": 233.823,
+      "eval_samples_per_second": 427.674,
+      "eval_steps_per_second": 13.365,
+      "step": 23500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.77477900189561e-05,
+      "loss": 2.4764,
+      "step": 24000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.2874138355255127,
+      "eval_runtime": 232.1465,
+      "eval_samples_per_second": 430.762,
+      "eval_steps_per_second": 13.461,
+      "step": 24000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.770086897768435e-05,
+      "loss": 2.4725,
+      "step": 24500
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.2882509231567383,
+      "eval_runtime": 233.6105,
+      "eval_samples_per_second": 428.063,
+      "eval_steps_per_second": 13.377,
+      "step": 24500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.7653947936412606e-05,
+      "loss": 2.4483,
+      "step": 25000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 2.2681896686553955,
+      "eval_runtime": 230.7325,
+      "eval_samples_per_second": 433.402,
+      "eval_steps_per_second": 13.544,
+      "step": 25000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.760702689514086e-05,
+      "loss": 2.4408,
+      "step": 25500
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.2558085918426514,
+      "eval_runtime": 231.5715,
+      "eval_samples_per_second": 431.832,
+      "eval_steps_per_second": 13.495,
+      "step": 25500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.756010585386911e-05,
+      "loss": 2.4344,
+      "step": 26000
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 2.2374377250671387,
+      "eval_runtime": 233.7469,
+      "eval_samples_per_second": 427.813,
+      "eval_steps_per_second": 13.369,
+      "step": 26000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.751318481259736e-05,
+      "loss": 2.4257,
+      "step": 26500
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.229942560195923,
+      "eval_runtime": 231.8416,
+      "eval_samples_per_second": 431.329,
+      "eval_steps_per_second": 13.479,
+      "step": 26500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.746626377132561e-05,
+      "loss": 2.3836,
+      "step": 27000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 2.2122104167938232,
+      "eval_runtime": 232.1939,
+      "eval_samples_per_second": 430.675,
+      "eval_steps_per_second": 13.459,
+      "step": 27000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.7419342730053865e-05,
+      "loss": 2.3901,
+      "step": 27500
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.1925010681152344,
+      "eval_runtime": 228.7288,
+      "eval_samples_per_second": 437.199,
+      "eval_steps_per_second": 13.662,
+      "step": 27500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.737242168878212e-05,
+      "loss": 2.3796,
+      "step": 28000
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 2.195193290710449,
+      "eval_runtime": 227.6921,
+      "eval_samples_per_second": 439.19,
+      "eval_steps_per_second": 13.725,
+      "step": 28000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.732550064751037e-05,
+      "loss": 2.3792,
+      "step": 28500
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.178008794784546,
+      "eval_runtime": 227.653,
+      "eval_samples_per_second": 439.265,
+      "eval_steps_per_second": 13.727,
+      "step": 28500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7278579606238625e-05,
+      "loss": 2.3715,
+      "step": 29000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 2.175689220428467,
+      "eval_runtime": 227.4287,
+      "eval_samples_per_second": 439.698,
+      "eval_steps_per_second": 13.741,
+      "step": 29000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.723165856496687e-05,
+      "loss": 2.387,
+      "step": 29500
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.17189884185791,
+      "eval_runtime": 227.5833,
+      "eval_samples_per_second": 439.399,
+      "eval_steps_per_second": 13.731,
+      "step": 29500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7184737523695125e-05,
+      "loss": 2.3689,
+      "step": 30000
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.1487345695495605,
+      "eval_runtime": 227.5611,
+      "eval_samples_per_second": 439.442,
+      "eval_steps_per_second": 13.733,
+      "step": 30000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.713781648242338e-05,
+      "loss": 2.3086,
+      "step": 30500
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.1480958461761475,
+      "eval_runtime": 227.4844,
+      "eval_samples_per_second": 439.591,
+      "eval_steps_per_second": 13.737,
+      "step": 30500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.709089544115163e-05,
+      "loss": 2.3212,
+      "step": 31000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.140958547592163,
+      "eval_runtime": 227.7332,
+      "eval_samples_per_second": 439.11,
+      "eval_steps_per_second": 13.722,
+      "step": 31000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.7043974399879885e-05,
+      "loss": 2.321,
+      "step": 31500
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 2.1231191158294678,
+      "eval_runtime": 227.7051,
+      "eval_samples_per_second": 439.164,
+      "eval_steps_per_second": 13.724,
+      "step": 31500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.699705335860813e-05,
+      "loss": 2.2855,
+      "step": 32000
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 2.1169111728668213,
+      "eval_runtime": 227.254,
+      "eval_samples_per_second": 440.036,
+      "eval_steps_per_second": 13.751,
+      "step": 32000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.6950132317336385e-05,
+      "loss": 2.2906,
+      "step": 32500
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 2.104003667831421,
+      "eval_runtime": 227.1649,
+      "eval_samples_per_second": 440.209,
+      "eval_steps_per_second": 13.757,
+      "step": 32500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.690321127606464e-05,
+      "loss": 2.2771,
+      "step": 33000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 2.0972249507904053,
+      "eval_runtime": 227.2504,
+      "eval_samples_per_second": 440.043,
+      "eval_steps_per_second": 13.751,
+      "step": 33000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.685629023479289e-05,
+      "loss": 2.2485,
+      "step": 33500
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 2.074732780456543,
+      "eval_runtime": 227.2974,
+      "eval_samples_per_second": 439.952,
+      "eval_steps_per_second": 13.749,
+      "step": 33500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.6809369193521144e-05,
+      "loss": 2.2783,
+      "step": 34000
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 2.077274799346924,
+      "eval_runtime": 227.3063,
+      "eval_samples_per_second": 439.935,
+      "eval_steps_per_second": 13.748,
+      "step": 34000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.67624481522494e-05,
+      "loss": 2.2668,
+      "step": 34500
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 2.075364828109741,
+      "eval_runtime": 227.2985,
+      "eval_samples_per_second": 439.95,
+      "eval_steps_per_second": 13.748,
+      "step": 34500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6715527110977644e-05,
+      "loss": 2.2458,
+      "step": 35000
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 2.060246467590332,
+      "eval_runtime": 227.0553,
+      "eval_samples_per_second": 440.421,
+      "eval_steps_per_second": 13.763,
+      "step": 35000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.66686060697059e-05,
+      "loss": 2.2365,
+      "step": 35500
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 2.046921968460083,
+      "eval_runtime": 227.3148,
+      "eval_samples_per_second": 439.919,
+      "eval_steps_per_second": 13.747,
+      "step": 35500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.662168502843415e-05,
+      "loss": 2.2652,
+      "step": 36000
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 2.049044370651245,
+      "eval_runtime": 227.288,
+      "eval_samples_per_second": 439.97,
+      "eval_steps_per_second": 13.749,
+      "step": 36000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6574763987162404e-05,
+      "loss": 2.2276,
+      "step": 36500
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 2.0362257957458496,
+      "eval_runtime": 227.1732,
+      "eval_samples_per_second": 440.193,
+      "eval_steps_per_second": 13.756,
+      "step": 36500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.652784294589066e-05,
+      "loss": 2.226,
+      "step": 37000
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 2.0287837982177734,
+      "eval_runtime": 227.1386,
+      "eval_samples_per_second": 440.26,
+      "eval_steps_per_second": 13.758,
+      "step": 37000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6480921904618904e-05,
+      "loss": 2.2358,
+      "step": 37500
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 2.016014337539673,
+      "eval_runtime": 227.3457,
+      "eval_samples_per_second": 439.859,
+      "eval_steps_per_second": 13.746,
+      "step": 37500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.643400086334716e-05,
+      "loss": 2.1975,
+      "step": 38000
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 2.0150697231292725,
+      "eval_runtime": 227.4193,
+      "eval_samples_per_second": 439.716,
+      "eval_steps_per_second": 13.741,
+      "step": 38000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.638707982207541e-05,
+      "loss": 2.1991,
+      "step": 38500
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 2.007798910140991,
+      "eval_runtime": 227.3815,
+      "eval_samples_per_second": 439.789,
+      "eval_steps_per_second": 13.743,
+      "step": 38500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6340158780803664e-05,
+      "loss": 2.1907,
+      "step": 39000
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 2.0014853477478027,
+      "eval_runtime": 227.3276,
+      "eval_samples_per_second": 439.894,
+      "eval_steps_per_second": 13.747,
+      "step": 39000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.629323773953192e-05,
+      "loss": 2.1551,
+      "step": 39500
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 2.0021867752075195,
+      "eval_runtime": 227.2631,
+      "eval_samples_per_second": 440.019,
+      "eval_steps_per_second": 13.751,
+      "step": 39500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.624631669826017e-05,
+      "loss": 2.1818,
+      "step": 40000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.9883029460906982,
+      "eval_runtime": 227.0668,
+      "eval_samples_per_second": 440.399,
+      "eval_steps_per_second": 13.762,
+      "step": 40000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.619939565698842e-05,
+      "loss": 2.1678,
+      "step": 40500
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.9864903688430786,
+      "eval_runtime": 226.6949,
+      "eval_samples_per_second": 441.122,
+      "eval_steps_per_second": 13.785,
+      "step": 40500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.615247461571667e-05,
+      "loss": 2.173,
+      "step": 41000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.9766247272491455,
+      "eval_runtime": 226.7708,
+      "eval_samples_per_second": 440.974,
+      "eval_steps_per_second": 13.78,
+      "step": 41000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6105553574444923e-05,
+      "loss": 2.1469,
+      "step": 41500
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.962270975112915,
+      "eval_runtime": 227.1551,
+      "eval_samples_per_second": 440.228,
+      "eval_steps_per_second": 13.757,
+      "step": 41500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.605863253317318e-05,
+      "loss": 2.1414,
+      "step": 42000
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 1.9634250402450562,
+      "eval_runtime": 227.1053,
+      "eval_samples_per_second": 440.324,
+      "eval_steps_per_second": 13.76,
+      "step": 42000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.601171149190143e-05,
+      "loss": 2.156,
+      "step": 42500
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.9548449516296387,
+      "eval_runtime": 226.8657,
+      "eval_samples_per_second": 440.789,
+      "eval_steps_per_second": 13.775,
+      "step": 42500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.596479045062968e-05,
+      "loss": 2.1155,
+      "step": 43000
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.9472512006759644,
+      "eval_runtime": 227.1058,
+      "eval_samples_per_second": 440.323,
+      "eval_steps_per_second": 13.76,
+      "step": 43000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.591786940935793e-05,
+      "loss": 2.114,
+      "step": 43500
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.943420171737671,
+      "eval_runtime": 227.1581,
+      "eval_samples_per_second": 440.222,
+      "eval_steps_per_second": 13.757,
+      "step": 43500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.587094836808618e-05,
+      "loss": 2.1146,
+      "step": 44000
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.9389326572418213,
+      "eval_runtime": 227.1632,
+      "eval_samples_per_second": 440.212,
+      "eval_steps_per_second": 13.757,
+      "step": 44000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5824027326814436e-05,
+      "loss": 2.1018,
+      "step": 44500
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.9317586421966553,
+      "eval_runtime": 227.2543,
+      "eval_samples_per_second": 440.036,
+      "eval_steps_per_second": 13.751,
+      "step": 44500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.577710628554269e-05,
+      "loss": 2.0955,
+      "step": 45000
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.9257296323776245,
+      "eval_runtime": 227.0518,
+      "eval_samples_per_second": 440.428,
+      "eval_steps_per_second": 13.763,
+      "step": 45000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.573018524427094e-05,
+      "loss": 2.0788,
+      "step": 45500
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.9265824556350708,
+      "eval_runtime": 226.9844,
+      "eval_samples_per_second": 440.559,
+      "eval_steps_per_second": 13.767,
+      "step": 45500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.5683264202999196e-05,
+      "loss": 2.0983,
+      "step": 46000
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.9132778644561768,
+      "eval_runtime": 226.9226,
+      "eval_samples_per_second": 440.679,
+      "eval_steps_per_second": 13.771,
+      "step": 46000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.563634316172745e-05,
+      "loss": 2.0877,
+      "step": 46500
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.9130772352218628,
+      "eval_runtime": 227.0233,
+      "eval_samples_per_second": 440.483,
+      "eval_steps_per_second": 13.765,
+      "step": 46500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.5589422120455696e-05,
+      "loss": 2.0831,
+      "step": 47000
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.9047877788543701,
+      "eval_runtime": 227.0434,
+      "eval_samples_per_second": 440.444,
+      "eval_steps_per_second": 13.764,
+      "step": 47000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.554250107918395e-05,
+      "loss": 2.0843,
+      "step": 47500
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.9129490852355957,
+      "eval_runtime": 226.7451,
+      "eval_samples_per_second": 441.024,
+      "eval_steps_per_second": 13.782,
+      "step": 47500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.54955800379122e-05,
+      "loss": 2.0583,
+      "step": 48000
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.8967323303222656,
+      "eval_runtime": 226.9943,
+      "eval_samples_per_second": 440.54,
+      "eval_steps_per_second": 13.767,
+      "step": 48000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.5448658996640456e-05,
+      "loss": 2.0814,
+      "step": 48500
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.8939955234527588,
+      "eval_runtime": 226.7956,
+      "eval_samples_per_second": 440.926,
+      "eval_steps_per_second": 13.779,
+      "step": 48500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.540173795536871e-05,
+      "loss": 2.0656,
+      "step": 49000
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.8772351741790771,
+      "eval_runtime": 226.9522,
+      "eval_samples_per_second": 440.621,
+      "eval_steps_per_second": 13.769,
+      "step": 49000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.535481691409696e-05,
+      "loss": 2.0654,
+      "step": 49500
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.8778467178344727,
+      "eval_runtime": 227.1023,
+      "eval_samples_per_second": 440.33,
+      "eval_steps_per_second": 13.76,
+      "step": 49500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.5307895872825216e-05,
+      "loss": 2.0695,
+      "step": 50000
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.877779245376587,
+      "eval_runtime": 226.8595,
+      "eval_samples_per_second": 440.801,
+      "eval_steps_per_second": 13.775,
+      "step": 50000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.526097483155346e-05,
+      "loss": 2.0571,
+      "step": 50500
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.878071904182434,
+      "eval_runtime": 226.8228,
+      "eval_samples_per_second": 440.873,
+      "eval_steps_per_second": 13.777,
+      "step": 50500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.5214053790281715e-05,
+      "loss": 2.0379,
+      "step": 51000
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.8661781549453735,
+      "eval_runtime": 226.9895,
+      "eval_samples_per_second": 440.549,
+      "eval_steps_per_second": 13.767,
+      "step": 51000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.516713274900997e-05,
+      "loss": 2.0439,
+      "step": 51500
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.8483119010925293,
+      "eval_runtime": 227.0054,
+      "eval_samples_per_second": 440.518,
+      "eval_steps_per_second": 13.766,
+      "step": 51500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.512021170773822e-05,
+      "loss": 2.0472,
+      "step": 52000
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.8448209762573242,
+      "eval_runtime": 226.9827,
+      "eval_samples_per_second": 440.562,
+      "eval_steps_per_second": 13.768,
+      "step": 52000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.5073290666466475e-05,
+      "loss": 2.0281,
+      "step": 52500
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 1.8498718738555908,
+      "eval_runtime": 227.1635,
+      "eval_samples_per_second": 440.212,
+      "eval_steps_per_second": 13.757,
+      "step": 52500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.502636962519473e-05,
+      "loss": 2.0077,
+      "step": 53000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.848965048789978,
+      "eval_runtime": 226.7848,
+      "eval_samples_per_second": 440.947,
+      "eval_steps_per_second": 13.78,
+      "step": 53000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.4979448583922975e-05,
+      "loss": 2.0407,
+      "step": 53500
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.8440462350845337,
+      "eval_runtime": 227.0578,
+      "eval_samples_per_second": 440.416,
+      "eval_steps_per_second": 13.763,
+      "step": 53500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.493252754265123e-05,
+      "loss": 2.0314,
+      "step": 54000
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.8299942016601562,
+      "eval_runtime": 226.8309,
+      "eval_samples_per_second": 440.857,
+      "eval_steps_per_second": 13.777,
+      "step": 54000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.488560650137948e-05,
+      "loss": 2.0178,
+      "step": 54500
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.8236579895019531,
+      "eval_runtime": 226.9072,
+      "eval_samples_per_second": 440.709,
+      "eval_steps_per_second": 13.772,
+      "step": 54500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.4838685460107735e-05,
+      "loss": 2.0133,
+      "step": 55000
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.826443076133728,
+      "eval_runtime": 226.9793,
+      "eval_samples_per_second": 440.569,
+      "eval_steps_per_second": 13.768,
+      "step": 55000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.479176441883599e-05,
+      "loss": 2.0298,
+      "step": 55500
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.816024661064148,
+      "eval_runtime": 226.7364,
+      "eval_samples_per_second": 441.041,
+      "eval_steps_per_second": 13.783,
+      "step": 55500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.474484337756424e-05,
+      "loss": 1.9824,
+      "step": 56000
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.8234096765518188,
+      "eval_runtime": 226.9723,
+      "eval_samples_per_second": 440.582,
+      "eval_steps_per_second": 13.768,
+      "step": 56000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.4697922336292495e-05,
+      "loss": 1.9908,
+      "step": 56500
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.8151246309280396,
+      "eval_runtime": 226.9337,
+      "eval_samples_per_second": 440.657,
+      "eval_steps_per_second": 13.771,
+      "step": 56500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.465100129502074e-05,
+      "loss": 1.9782,
+      "step": 57000
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 1.808193564414978,
+      "eval_runtime": 226.7205,
+      "eval_samples_per_second": 441.072,
+      "eval_steps_per_second": 13.783,
+      "step": 57000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.4604080253748995e-05,
+      "loss": 1.9985,
+      "step": 57500
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.7979633808135986,
+      "eval_runtime": 226.818,
+      "eval_samples_per_second": 440.882,
+      "eval_steps_per_second": 13.778,
+      "step": 57500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.455715921247725e-05,
+      "loss": 1.978,
+      "step": 58000
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.7955108880996704,
+      "eval_runtime": 226.7306,
+      "eval_samples_per_second": 441.052,
+      "eval_steps_per_second": 13.783,
+      "step": 58000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.45102381712055e-05,
+      "loss": 1.9746,
+      "step": 58500
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.794771671295166,
+      "eval_runtime": 226.9956,
+      "eval_samples_per_second": 440.537,
+      "eval_steps_per_second": 13.767,
+      "step": 58500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.4463317129933754e-05,
+      "loss": 1.9838,
+      "step": 59000
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.7896528244018555,
+      "eval_runtime": 226.9957,
+      "eval_samples_per_second": 440.537,
+      "eval_steps_per_second": 13.767,
+      "step": 59000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4416396088662e-05,
+      "loss": 1.944,
+      "step": 59500
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.7889646291732788,
+      "eval_runtime": 227.1746,
+      "eval_samples_per_second": 440.19,
+      "eval_steps_per_second": 13.756,
+      "step": 59500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.4369475047390254e-05,
+      "loss": 1.965,
+      "step": 60000
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 1.7891310453414917,
+      "eval_runtime": 227.2101,
+      "eval_samples_per_second": 440.121,
+      "eval_steps_per_second": 13.754,
+      "step": 60000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.432255400611851e-05,
+      "loss": 1.9501,
+      "step": 60500
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 1.7808892726898193,
+      "eval_runtime": 226.9599,
+      "eval_samples_per_second": 440.607,
+      "eval_steps_per_second": 13.769,
+      "step": 60500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.427563296484676e-05,
+      "loss": 1.9541,
+      "step": 61000
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 1.7755423784255981,
+      "eval_runtime": 226.9182,
+      "eval_samples_per_second": 440.688,
+      "eval_steps_per_second": 13.771,
+      "step": 61000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.4228711923575014e-05,
+      "loss": 1.9483,
+      "step": 61500
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.7737643718719482,
+      "eval_runtime": 226.5336,
+      "eval_samples_per_second": 441.436,
+      "eval_steps_per_second": 13.795,
+      "step": 61500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.418179088230327e-05,
+      "loss": 1.9568,
+      "step": 62000
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.7714828252792358,
+      "eval_runtime": 226.801,
+      "eval_samples_per_second": 440.915,
+      "eval_steps_per_second": 13.779,
+      "step": 62000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.4134869841031514e-05,
+      "loss": 1.9418,
+      "step": 62500
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.7680498361587524,
+      "eval_runtime": 226.5426,
+      "eval_samples_per_second": 441.418,
+      "eval_steps_per_second": 13.794,
+      "step": 62500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.408794879975977e-05,
+      "loss": 1.946,
+      "step": 63000
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.756327509880066,
+      "eval_runtime": 226.7739,
+      "eval_samples_per_second": 440.968,
+      "eval_steps_per_second": 13.78,
+      "step": 63000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.404102775848802e-05,
+      "loss": 1.9331,
+      "step": 63500
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.7607768774032593,
+      "eval_runtime": 226.7955,
+      "eval_samples_per_second": 440.926,
+      "eval_steps_per_second": 13.779,
+      "step": 63500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.3994106717216274e-05,
+      "loss": 1.9287,
+      "step": 64000
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.7537603378295898,
+      "eval_runtime": 226.7832,
+      "eval_samples_per_second": 440.95,
+      "eval_steps_per_second": 13.78,
+      "step": 64000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.394718567594453e-05,
+      "loss": 1.919,
+      "step": 64500
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.7554343938827515,
+      "eval_runtime": 226.8184,
+      "eval_samples_per_second": 440.881,
+      "eval_steps_per_second": 13.778,
+      "step": 64500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.390026463467278e-05,
+      "loss": 1.9506,
+      "step": 65000
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.744460105895996,
+      "eval_runtime": 226.7808,
+      "eval_samples_per_second": 440.954,
+      "eval_steps_per_second": 13.78,
+      "step": 65000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.385334359340103e-05,
+      "loss": 1.9358,
+      "step": 65500
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.7381893396377563,
+      "eval_runtime": 226.843,
+      "eval_samples_per_second": 440.833,
+      "eval_steps_per_second": 13.776,
+      "step": 65500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.380642255212928e-05,
+      "loss": 1.9002,
+      "step": 66000
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.7369801998138428,
+      "eval_runtime": 226.5756,
+      "eval_samples_per_second": 441.354,
+      "eval_steps_per_second": 13.792,
+      "step": 66000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.3759501510857533e-05,
+      "loss": 1.9122,
+      "step": 66500
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.7368305921554565,
+      "eval_runtime": 226.767,
+      "eval_samples_per_second": 440.981,
+      "eval_steps_per_second": 13.781,
+      "step": 66500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.371258046958579e-05,
+      "loss": 1.9225,
+      "step": 67000
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.7325348854064941,
+      "eval_runtime": 226.9377,
+      "eval_samples_per_second": 440.65,
+      "eval_steps_per_second": 13.77,
+      "step": 67000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.366565942831404e-05,
+      "loss": 1.9213,
+      "step": 67500
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.7244173288345337,
+      "eval_runtime": 226.9908,
+      "eval_samples_per_second": 440.547,
+      "eval_steps_per_second": 13.767,
+      "step": 67500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.3618738387042287e-05,
+      "loss": 1.9084,
+      "step": 68000
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.725306510925293,
+      "eval_runtime": 226.9071,
+      "eval_samples_per_second": 440.709,
+      "eval_steps_per_second": 13.772,
+      "step": 68000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.357181734577054e-05,
+      "loss": 1.9056,
+      "step": 68500
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.716600775718689,
+      "eval_runtime": 226.8672,
+      "eval_samples_per_second": 440.786,
+      "eval_steps_per_second": 13.775,
+      "step": 68500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.352489630449879e-05,
+      "loss": 1.9175,
+      "step": 69000
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 1.7228271961212158,
+      "eval_runtime": 226.6113,
+      "eval_samples_per_second": 441.284,
+      "eval_steps_per_second": 13.79,
+      "step": 69000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.3477975263227046e-05,
+      "loss": 1.8865,
+      "step": 69500
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 1.7209404706954956,
+      "eval_runtime": 226.7014,
+      "eval_samples_per_second": 441.109,
+      "eval_steps_per_second": 13.785,
+      "step": 69500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.34310542219553e-05,
+      "loss": 1.8866,
+      "step": 70000
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.7054914236068726,
+      "eval_runtime": 226.4302,
+      "eval_samples_per_second": 441.637,
+      "eval_steps_per_second": 13.801,
+      "step": 70000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.338413318068355e-05,
+      "loss": 1.8733,
+      "step": 70500
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.712384819984436,
+      "eval_runtime": 226.2537,
+      "eval_samples_per_second": 441.982,
+      "eval_steps_per_second": 13.812,
+      "step": 70500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.33372121394118e-05,
+      "loss": 1.9032,
+      "step": 71000
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.6990084648132324,
+      "eval_runtime": 225.9511,
+      "eval_samples_per_second": 442.574,
+      "eval_steps_per_second": 13.83,
+      "step": 71000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.329029109814005e-05,
+      "loss": 1.8713,
+      "step": 71500
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.6956583261489868,
+      "eval_runtime": 226.2467,
+      "eval_samples_per_second": 441.995,
+      "eval_steps_per_second": 13.812,
+      "step": 71500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.3243370056868306e-05,
+      "loss": 1.8766,
+      "step": 72000
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.6951245069503784,
+      "eval_runtime": 226.2025,
+      "eval_samples_per_second": 442.082,
+      "eval_steps_per_second": 13.815,
+      "step": 72000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.319644901559656e-05,
+      "loss": 1.9066,
+      "step": 72500
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.6982998847961426,
+      "eval_runtime": 226.163,
+      "eval_samples_per_second": 442.159,
+      "eval_steps_per_second": 13.817,
+      "step": 72500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.314952797432481e-05,
+      "loss": 1.889,
+      "step": 73000
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.7046481370925903,
+      "eval_runtime": 226.2165,
+      "eval_samples_per_second": 442.054,
+      "eval_steps_per_second": 13.814,
+      "step": 73000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.310260693305306e-05,
+      "loss": 1.8583,
+      "step": 73500
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.6921665668487549,
+      "eval_runtime": 226.1668,
+      "eval_samples_per_second": 442.152,
+      "eval_steps_per_second": 13.817,
+      "step": 73500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.305568589178131e-05,
+      "loss": 1.846,
+      "step": 74000
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.6901285648345947,
+      "eval_runtime": 225.9398,
+      "eval_samples_per_second": 442.596,
+      "eval_steps_per_second": 13.831,
+      "step": 74000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.3008764850509566e-05,
+      "loss": 1.8748,
+      "step": 74500
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.6798553466796875,
+      "eval_runtime": 226.2378,
+      "eval_samples_per_second": 442.013,
+      "eval_steps_per_second": 13.813,
+      "step": 74500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.296184380923782e-05,
+      "loss": 1.8596,
+      "step": 75000
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 1.681208848953247,
+      "eval_runtime": 226.5798,
+      "eval_samples_per_second": 441.346,
+      "eval_steps_per_second": 13.792,
+      "step": 75000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.291492276796607e-05,
+      "loss": 1.843,
+      "step": 75500
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.6777198314666748,
+      "eval_runtime": 226.8856,
+      "eval_samples_per_second": 440.751,
+      "eval_steps_per_second": 13.773,
+      "step": 75500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.2868001726694326e-05,
+      "loss": 1.8458,
+      "step": 76000
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.6699376106262207,
+      "eval_runtime": 226.7938,
+      "eval_samples_per_second": 440.929,
+      "eval_steps_per_second": 13.779,
+      "step": 76000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.282108068542257e-05,
+      "loss": 1.8544,
+      "step": 76500
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.6637232303619385,
+      "eval_runtime": 226.7851,
+      "eval_samples_per_second": 440.946,
+      "eval_steps_per_second": 13.78,
+      "step": 76500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.2774159644150825e-05,
+      "loss": 1.8533,
+      "step": 77000
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.6666810512542725,
+      "eval_runtime": 226.4743,
+      "eval_samples_per_second": 441.551,
+      "eval_steps_per_second": 13.798,
+      "step": 77000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.272723860287908e-05,
+      "loss": 1.8645,
+      "step": 77500
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 1.66965651512146,
+      "eval_runtime": 226.5459,
+      "eval_samples_per_second": 441.412,
+      "eval_steps_per_second": 13.794,
+      "step": 77500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.268031756160733e-05,
+      "loss": 1.8535,
+      "step": 78000
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 1.665113091468811,
+      "eval_runtime": 226.5387,
+      "eval_samples_per_second": 441.426,
+      "eval_steps_per_second": 13.795,
+      "step": 78000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.2633396520335585e-05,
+      "loss": 1.8425,
+      "step": 78500
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.655056118965149,
+      "eval_runtime": 226.5932,
+      "eval_samples_per_second": 441.319,
+      "eval_steps_per_second": 13.791,
+      "step": 78500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.258647547906383e-05,
+      "loss": 1.8642,
+      "step": 79000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.6554076671600342,
+      "eval_runtime": 226.5747,
+      "eval_samples_per_second": 441.356,
+      "eval_steps_per_second": 13.792,
+      "step": 79000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2539554437792085e-05,
+      "loss": 1.8259,
+      "step": 79500
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.6575286388397217,
+      "eval_runtime": 226.3838,
+      "eval_samples_per_second": 441.728,
+      "eval_steps_per_second": 13.804,
+      "step": 79500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.249263339652034e-05,
+      "loss": 1.8199,
+      "step": 80000
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.6621437072753906,
+      "eval_runtime": 226.5957,
+      "eval_samples_per_second": 441.315,
+      "eval_steps_per_second": 13.791,
+      "step": 80000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.244571235524859e-05,
+      "loss": 1.8384,
+      "step": 80500
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 1.6512243747711182,
+      "eval_runtime": 226.3597,
+      "eval_samples_per_second": 441.775,
+      "eval_steps_per_second": 13.805,
+      "step": 80500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.2398791313976845e-05,
+      "loss": 1.8349,
+      "step": 81000
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 1.6566110849380493,
+      "eval_runtime": 226.3468,
+      "eval_samples_per_second": 441.8,
+      "eval_steps_per_second": 13.806,
+      "step": 81000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.23518702727051e-05,
+      "loss": 1.8342,
+      "step": 81500
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 1.6478888988494873,
+      "eval_runtime": 226.3371,
+      "eval_samples_per_second": 441.819,
+      "eval_steps_per_second": 13.807,
+      "step": 81500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.2304949231433345e-05,
+      "loss": 1.8508,
+      "step": 82000
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.6489554643630981,
+      "eval_runtime": 226.3981,
+      "eval_samples_per_second": 441.7,
+      "eval_steps_per_second": 13.803,
+      "step": 82000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.22580281901616e-05,
+      "loss": 1.8135,
+      "step": 82500
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.6381429433822632,
+      "eval_runtime": 226.3351,
+      "eval_samples_per_second": 441.823,
+      "eval_steps_per_second": 13.807,
+      "step": 82500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.221110714888985e-05,
+      "loss": 1.8332,
+      "step": 83000
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 1.6337202787399292,
+      "eval_runtime": 226.073,
+      "eval_samples_per_second": 442.335,
+      "eval_steps_per_second": 13.823,
+      "step": 83000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.2164186107618104e-05,
+      "loss": 1.8315,
+      "step": 83500
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 1.637961506843567,
+      "eval_runtime": 226.4701,
+      "eval_samples_per_second": 441.559,
+      "eval_steps_per_second": 13.799,
+      "step": 83500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.211726506634636e-05,
+      "loss": 1.8363,
+      "step": 84000
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.6368101835250854,
+      "eval_runtime": 226.1996,
+      "eval_samples_per_second": 442.087,
+      "eval_steps_per_second": 13.815,
+      "step": 84000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2070344025074604e-05,
+      "loss": 1.8248,
+      "step": 84500
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.626143455505371,
+      "eval_runtime": 226.4446,
+      "eval_samples_per_second": 441.609,
+      "eval_steps_per_second": 13.8,
+      "step": 84500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.202342298380286e-05,
+      "loss": 1.8045,
+      "step": 85000
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.6303765773773193,
+      "eval_runtime": 226.3791,
+      "eval_samples_per_second": 441.737,
+      "eval_steps_per_second": 13.804,
+      "step": 85000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.197650194253111e-05,
+      "loss": 1.8172,
+      "step": 85500
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.6250081062316895,
+      "eval_runtime": 226.1683,
+      "eval_samples_per_second": 442.149,
+      "eval_steps_per_second": 13.817,
+      "step": 85500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.1929580901259364e-05,
+      "loss": 1.8161,
+      "step": 86000
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.615513801574707,
+      "eval_runtime": 226.0153,
+      "eval_samples_per_second": 442.448,
+      "eval_steps_per_second": 13.826,
+      "step": 86000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.188265985998762e-05,
+      "loss": 1.7852,
+      "step": 86500
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.6129599809646606,
+      "eval_runtime": 225.9925,
+      "eval_samples_per_second": 442.493,
+      "eval_steps_per_second": 13.828,
+      "step": 86500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.183573881871587e-05,
+      "loss": 1.7935,
+      "step": 87000
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 1.6187821626663208,
+      "eval_runtime": 226.2683,
+      "eval_samples_per_second": 441.953,
+      "eval_steps_per_second": 13.811,
+      "step": 87000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.178881777744412e-05,
+      "loss": 1.791,
+      "step": 87500
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 1.6135245561599731,
+      "eval_runtime": 226.3101,
+      "eval_samples_per_second": 441.872,
+      "eval_steps_per_second": 13.808,
+      "step": 87500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.174189673617237e-05,
+      "loss": 1.8019,
+      "step": 88000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.6148468255996704,
+      "eval_runtime": 225.9824,
+      "eval_samples_per_second": 442.512,
+      "eval_steps_per_second": 13.829,
+      "step": 88000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.1694975694900624e-05,
+      "loss": 1.7808,
+      "step": 88500
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.6106659173965454,
+      "eval_runtime": 226.0203,
+      "eval_samples_per_second": 442.438,
+      "eval_steps_per_second": 13.826,
+      "step": 88500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.164805465362888e-05,
+      "loss": 1.7844,
+      "step": 89000
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.6136785745620728,
+      "eval_runtime": 226.2859,
+      "eval_samples_per_second": 441.919,
+      "eval_steps_per_second": 13.81,
+      "step": 89000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.160113361235713e-05,
+      "loss": 1.774,
+      "step": 89500
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.6069365739822388,
+      "eval_runtime": 226.2313,
+      "eval_samples_per_second": 442.025,
+      "eval_steps_per_second": 13.813,
+      "step": 89500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.155421257108538e-05,
+      "loss": 1.7877,
+      "step": 90000
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.6087855100631714,
+      "eval_runtime": 226.3039,
+      "eval_samples_per_second": 441.884,
+      "eval_steps_per_second": 13.809,
+      "step": 90000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.150729152981363e-05,
+      "loss": 1.7757,
+      "step": 90500
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 1.590334177017212,
+      "eval_runtime": 226.2681,
+      "eval_samples_per_second": 441.954,
+      "eval_steps_per_second": 13.811,
+      "step": 90500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.1460370488541883e-05,
+      "loss": 1.7731,
+      "step": 91000
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 1.5941905975341797,
+      "eval_runtime": 226.0068,
+      "eval_samples_per_second": 442.465,
+      "eval_steps_per_second": 13.827,
+      "step": 91000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.141344944727014e-05,
+      "loss": 1.7639,
+      "step": 91500
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 1.6017308235168457,
+      "eval_runtime": 226.0481,
+      "eval_samples_per_second": 442.384,
+      "eval_steps_per_second": 13.824,
+      "step": 91500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.136652840599839e-05,
+      "loss": 1.7889,
+      "step": 92000
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 1.592837929725647,
+      "eval_runtime": 226.2492,
+      "eval_samples_per_second": 441.991,
+      "eval_steps_per_second": 13.812,
+      "step": 92000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.131960736472664e-05,
+      "loss": 1.7788,
+      "step": 92500
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 1.5879290103912354,
+      "eval_runtime": 226.1946,
+      "eval_samples_per_second": 442.097,
+      "eval_steps_per_second": 13.816,
+      "step": 92500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.127268632345489e-05,
+      "loss": 1.7976,
+      "step": 93000
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 1.5884945392608643,
+      "eval_runtime": 226.1587,
+      "eval_samples_per_second": 442.167,
+      "eval_steps_per_second": 13.818,
+      "step": 93000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.122576528218314e-05,
+      "loss": 1.7698,
+      "step": 93500
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.5887948274612427,
+      "eval_runtime": 226.2905,
+      "eval_samples_per_second": 441.91,
+      "eval_steps_per_second": 13.81,
+      "step": 93500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.1178844240911396e-05,
+      "loss": 1.7684,
+      "step": 94000
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.576420783996582,
+      "eval_runtime": 226.1872,
+      "eval_samples_per_second": 442.112,
+      "eval_steps_per_second": 13.816,
+      "step": 94000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.113192319963965e-05,
+      "loss": 1.7679,
+      "step": 94500
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.5760667324066162,
+      "eval_runtime": 226.1526,
+      "eval_samples_per_second": 442.179,
+      "eval_steps_per_second": 13.818,
+      "step": 94500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.10850021583679e-05,
+      "loss": 1.7532,
+      "step": 95000
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.576305627822876,
+      "eval_runtime": 226.1826,
+      "eval_samples_per_second": 442.121,
+      "eval_steps_per_second": 13.816,
+      "step": 95000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.103808111709615e-05,
+      "loss": 1.765,
+      "step": 95500
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.580979347229004,
+      "eval_runtime": 226.2119,
+      "eval_samples_per_second": 442.063,
+      "eval_steps_per_second": 13.814,
+      "step": 95500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.09911600758244e-05,
+      "loss": 1.7429,
+      "step": 96000
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 1.5754783153533936,
+      "eval_runtime": 226.2117,
+      "eval_samples_per_second": 442.064,
+      "eval_steps_per_second": 13.814,
+      "step": 96000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.0944239034552656e-05,
+      "loss": 1.7522,
+      "step": 96500
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.5753768682479858,
+      "eval_runtime": 225.941,
+      "eval_samples_per_second": 442.593,
+      "eval_steps_per_second": 13.831,
+      "step": 96500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.089731799328091e-05,
+      "loss": 1.7401,
+      "step": 97000
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.5701993703842163,
+      "eval_runtime": 225.8905,
+      "eval_samples_per_second": 442.692,
+      "eval_steps_per_second": 13.834,
+      "step": 97000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.085039695200916e-05,
+      "loss": 1.7533,
+      "step": 97500
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.5690381526947021,
+      "eval_runtime": 226.1613,
+      "eval_samples_per_second": 442.162,
+      "eval_steps_per_second": 13.818,
+      "step": 97500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.0803475910737416e-05,
+      "loss": 1.7558,
+      "step": 98000
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.564376711845398,
+      "eval_runtime": 226.2821,
+      "eval_samples_per_second": 441.926,
+      "eval_steps_per_second": 13.81,
+      "step": 98000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.075655486946566e-05,
+      "loss": 1.7256,
+      "step": 98500
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.562330722808838,
+      "eval_runtime": 226.2172,
+      "eval_samples_per_second": 442.053,
+      "eval_steps_per_second": 13.814,
+      "step": 98500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.0709633828193916e-05,
+      "loss": 1.7487,
+      "step": 99000
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 1.5604215860366821,
+      "eval_runtime": 225.9865,
+      "eval_samples_per_second": 442.504,
+      "eval_steps_per_second": 13.828,
+      "step": 99000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.066271278692217e-05,
+      "loss": 1.748,
+      "step": 99500
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 1.55665922164917,
+      "eval_runtime": 226.1873,
+      "eval_samples_per_second": 442.111,
+      "eval_steps_per_second": 13.816,
+      "step": 99500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.061579174565042e-05,
+      "loss": 1.7337,
+      "step": 100000
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.5678207874298096,
+      "eval_runtime": 226.1956,
+      "eval_samples_per_second": 442.095,
+      "eval_steps_per_second": 13.815,
+      "step": 100000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.0568870704378675e-05,
+      "loss": 1.7551,
+      "step": 100500
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.5610854625701904,
+      "eval_runtime": 226.1994,
+      "eval_samples_per_second": 442.088,
+      "eval_steps_per_second": 13.815,
+      "step": 100500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.052194966310692e-05,
+      "loss": 1.7394,
+      "step": 101000
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.5528591871261597,
+      "eval_runtime": 226.4258,
+      "eval_samples_per_second": 441.646,
+      "eval_steps_per_second": 13.801,
+      "step": 101000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.0475028621835175e-05,
+      "loss": 1.7447,
+      "step": 101500
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.5571315288543701,
+      "eval_runtime": 226.3464,
+      "eval_samples_per_second": 441.801,
+      "eval_steps_per_second": 13.806,
+      "step": 101500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.042810758056343e-05,
+      "loss": 1.7411,
+      "step": 102000
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.5496081113815308,
+      "eval_runtime": 226.2985,
+      "eval_samples_per_second": 441.894,
+      "eval_steps_per_second": 13.809,
+      "step": 102000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.038118653929168e-05,
+      "loss": 1.7496,
+      "step": 102500
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.551464319229126,
+      "eval_runtime": 226.3296,
+      "eval_samples_per_second": 441.834,
+      "eval_steps_per_second": 13.807,
+      "step": 102500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.0334265498019935e-05,
+      "loss": 1.7381,
+      "step": 103000
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 1.5396504402160645,
+      "eval_runtime": 225.928,
+      "eval_samples_per_second": 442.619,
+      "eval_steps_per_second": 13.832,
+      "step": 103000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.028734445674819e-05,
+      "loss": 1.7513,
+      "step": 103500
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 1.5501888990402222,
+      "eval_runtime": 226.1976,
+      "eval_samples_per_second": 442.091,
+      "eval_steps_per_second": 13.815,
+      "step": 103500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.0240423415476435e-05,
+      "loss": 1.7126,
+      "step": 104000
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.5375173091888428,
+      "eval_runtime": 226.1628,
+      "eval_samples_per_second": 442.159,
+      "eval_steps_per_second": 13.817,
+      "step": 104000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.019350237420469e-05,
+      "loss": 1.7644,
+      "step": 104500
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.5411220788955688,
+      "eval_runtime": 226.1986,
+      "eval_samples_per_second": 442.089,
+      "eval_steps_per_second": 13.815,
+      "step": 104500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.014658133293294e-05,
+      "loss": 1.7142,
+      "step": 105000
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.5340408086776733,
+      "eval_runtime": 226.2255,
+      "eval_samples_per_second": 442.037,
+      "eval_steps_per_second": 13.814,
+      "step": 105000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.0099660291661195e-05,
+      "loss": 1.7053,
+      "step": 105500
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.5434354543685913,
+      "eval_runtime": 226.2654,
+      "eval_samples_per_second": 441.959,
+      "eval_steps_per_second": 13.811,
+      "step": 105500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.005273925038945e-05,
+      "loss": 1.7352,
+      "step": 106000
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.5321134328842163,
+      "eval_runtime": 226.0566,
+      "eval_samples_per_second": 442.367,
+      "eval_steps_per_second": 13.824,
+      "step": 106000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.00058182091177e-05,
+      "loss": 1.7152,
+      "step": 106500
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.5397337675094604,
+      "eval_runtime": 226.363,
+      "eval_samples_per_second": 441.768,
+      "eval_steps_per_second": 13.805,
+      "step": 106500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995889716784595e-05,
+      "loss": 1.6927,
+      "step": 107000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.542881727218628,
+      "eval_runtime": 226.3007,
+      "eval_samples_per_second": 441.89,
+      "eval_steps_per_second": 13.809,
+      "step": 107000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.99119761265742e-05,
+      "loss": 1.7013,
+      "step": 107500
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.528303861618042,
+      "eval_runtime": 226.2289,
+      "eval_samples_per_second": 442.03,
+      "eval_steps_per_second": 13.813,
+      "step": 107500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9865055085302454e-05,
+      "loss": 1.7318,
+      "step": 108000
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.5268659591674805,
+      "eval_runtime": 226.1768,
+      "eval_samples_per_second": 442.132,
+      "eval_steps_per_second": 13.817,
+      "step": 108000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.981813404403071e-05,
+      "loss": 1.6903,
+      "step": 108500
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.528219223022461,
+      "eval_runtime": 226.4877,
+      "eval_samples_per_second": 441.525,
+      "eval_steps_per_second": 13.798,
+      "step": 108500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.977121300275896e-05,
+      "loss": 1.7105,
+      "step": 109000
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.5189323425292969,
+      "eval_runtime": 226.1979,
+      "eval_samples_per_second": 442.091,
+      "eval_steps_per_second": 13.815,
+      "step": 109000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.972429196148721e-05,
+      "loss": 1.6979,
+      "step": 109500
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 1.5255582332611084,
+      "eval_runtime": 226.4286,
+      "eval_samples_per_second": 441.64,
+      "eval_steps_per_second": 13.801,
+      "step": 109500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.967737092021546e-05,
+      "loss": 1.6922,
+      "step": 110000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 1.5171970129013062,
+      "eval_runtime": 226.4396,
+      "eval_samples_per_second": 441.619,
+      "eval_steps_per_second": 13.801,
+      "step": 110000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.9630449878943714e-05,
+      "loss": 1.7161,
+      "step": 110500
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.517843246459961,
+      "eval_runtime": 226.4686,
+      "eval_samples_per_second": 441.562,
+      "eval_steps_per_second": 13.799,
+      "step": 110500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.958352883767197e-05,
+      "loss": 1.719,
+      "step": 111000
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.5142602920532227,
+      "eval_runtime": 226.7344,
+      "eval_samples_per_second": 441.045,
+      "eval_steps_per_second": 13.783,
+      "step": 111000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.953660779640022e-05,
+      "loss": 1.6857,
+      "step": 111500
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.5159624814987183,
+      "eval_runtime": 226.7078,
+      "eval_samples_per_second": 441.096,
+      "eval_steps_per_second": 13.784,
+      "step": 111500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.9489686755128474e-05,
+      "loss": 1.6983,
+      "step": 112000
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.510577917098999,
+      "eval_runtime": 226.2373,
+      "eval_samples_per_second": 442.014,
+      "eval_steps_per_second": 13.813,
+      "step": 112000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.944276571385672e-05,
+      "loss": 1.6838,
+      "step": 112500
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.5174448490142822,
+      "eval_runtime": 226.647,
+      "eval_samples_per_second": 441.215,
+      "eval_steps_per_second": 13.788,
+      "step": 112500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.9395844672584974e-05,
+      "loss": 1.6936,
+      "step": 113000
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.5134094953536987,
+      "eval_runtime": 226.7825,
+      "eval_samples_per_second": 440.951,
+      "eval_steps_per_second": 13.78,
+      "step": 113000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.934892363131323e-05,
+      "loss": 1.675,
+      "step": 113500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.5090696811676025,
+      "eval_runtime": 226.8241,
+      "eval_samples_per_second": 440.87,
+      "eval_steps_per_second": 13.777,
+      "step": 113500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.930200259004148e-05,
+      "loss": 1.6706,
+      "step": 114000
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.5159416198730469,
+      "eval_runtime": 226.8746,
+      "eval_samples_per_second": 440.772,
+      "eval_steps_per_second": 13.774,
+      "step": 114000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.9255081548769734e-05,
+      "loss": 1.6763,
+      "step": 114500
+    },
+    {
+      "epoch": 1.07,
+      "eval_loss": 1.4984365701675415,
+      "eval_runtime": 226.8627,
+      "eval_samples_per_second": 440.795,
+      "eval_steps_per_second": 13.775,
+      "step": 114500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.920816050749798e-05,
+      "loss": 1.6888,
+      "step": 115000
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.504223108291626,
+      "eval_runtime": 228.3239,
+      "eval_samples_per_second": 437.974,
+      "eval_steps_per_second": 13.687,
+      "step": 115000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.916123946622623e-05,
+      "loss": 1.6627,
+      "step": 115500
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.5011541843414307,
+      "eval_runtime": 226.9808,
+      "eval_samples_per_second": 440.566,
+      "eval_steps_per_second": 13.768,
+      "step": 115500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.911431842495449e-05,
+      "loss": 1.6739,
+      "step": 116000
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.5042829513549805,
+      "eval_runtime": 227.232,
+      "eval_samples_per_second": 440.079,
+      "eval_steps_per_second": 13.752,
+      "step": 116000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.906739738368274e-05,
+      "loss": 1.6465,
+      "step": 116500
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.496393084526062,
+      "eval_runtime": 227.0422,
+      "eval_samples_per_second": 440.447,
+      "eval_steps_per_second": 13.764,
+      "step": 116500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.902047634241099e-05,
+      "loss": 1.681,
+      "step": 117000
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 1.501774549484253,
+      "eval_runtime": 226.8713,
+      "eval_samples_per_second": 440.779,
+      "eval_steps_per_second": 13.774,
+      "step": 117000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.8973555301139247e-05,
+      "loss": 1.6674,
+      "step": 117500
+    },
+    {
+      "epoch": 1.1,
+      "eval_loss": 1.492223858833313,
+      "eval_runtime": 227.2955,
+      "eval_samples_per_second": 439.956,
+      "eval_steps_per_second": 13.749,
+      "step": 117500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.892663425986749e-05,
+      "loss": 1.6808,
+      "step": 118000
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 1.5016900300979614,
+      "eval_runtime": 227.0887,
+      "eval_samples_per_second": 440.356,
+      "eval_steps_per_second": 13.761,
+      "step": 118000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.8879713218595746e-05,
+      "loss": 1.6687,
+      "step": 118500
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 1.4928853511810303,
+      "eval_runtime": 235.0541,
+      "eval_samples_per_second": 425.434,
+      "eval_steps_per_second": 13.295,
+      "step": 118500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.8832792177324e-05,
+      "loss": 1.6711,
+      "step": 119000
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.4951835870742798,
+      "eval_runtime": 231.8391,
+      "eval_samples_per_second": 431.334,
+      "eval_steps_per_second": 13.479,
+      "step": 119000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.878587113605225e-05,
+      "loss": 1.6629,
+      "step": 119500
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.49068284034729,
+      "eval_runtime": 232.7614,
+      "eval_samples_per_second": 429.624,
+      "eval_steps_per_second": 13.426,
+      "step": 119500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.8738950094780506e-05,
+      "loss": 1.6762,
+      "step": 120000
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.497752070426941,
+      "eval_runtime": 226.4934,
+      "eval_samples_per_second": 441.514,
+      "eval_steps_per_second": 13.797,
+      "step": 120000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.869202905350875e-05,
+      "loss": 1.6806,
+      "step": 120500
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.5004570484161377,
+      "eval_runtime": 231.6639,
+      "eval_samples_per_second": 431.66,
+      "eval_steps_per_second": 13.489,
+      "step": 120500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.8645108012237006e-05,
+      "loss": 1.6322,
+      "step": 121000
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 1.491471767425537,
+      "eval_runtime": 223.6465,
+      "eval_samples_per_second": 447.134,
+      "eval_steps_per_second": 13.973,
+      "step": 121000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.859818697096526e-05,
+      "loss": 1.6823,
+      "step": 121500
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 1.4876571893692017,
+      "eval_runtime": 223.7062,
+      "eval_samples_per_second": 447.015,
+      "eval_steps_per_second": 13.969,
+      "step": 121500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.855126592969351e-05,
+      "loss": 1.6659,
+      "step": 122000
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 1.4812325239181519,
+      "eval_runtime": 223.6584,
+      "eval_samples_per_second": 447.11,
+      "eval_steps_per_second": 13.972,
+      "step": 122000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.8504344888421766e-05,
+      "loss": 1.6801,
+      "step": 122500
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.4771298170089722,
+      "eval_runtime": 223.7401,
+      "eval_samples_per_second": 446.947,
+      "eval_steps_per_second": 13.967,
+      "step": 122500
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.845742384715002e-05,
+      "loss": 1.6685,
+      "step": 123000
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.473673939704895,
+      "eval_runtime": 223.7681,
+      "eval_samples_per_second": 446.891,
+      "eval_steps_per_second": 13.965,
+      "step": 123000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.8410502805878266e-05,
+      "loss": 1.6686,
+      "step": 123500
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.4780343770980835,
+      "eval_runtime": 223.8692,
+      "eval_samples_per_second": 446.689,
+      "eval_steps_per_second": 13.959,
+      "step": 123500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.836358176460652e-05,
+      "loss": 1.6603,
+      "step": 124000
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.478870153427124,
+      "eval_runtime": 237.4256,
+      "eval_samples_per_second": 421.185,
+      "eval_steps_per_second": 13.162,
+      "step": 124000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.831666072333477e-05,
+      "loss": 1.6402,
+      "step": 124500
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 1.4783562421798706,
+      "eval_runtime": 236.4563,
+      "eval_samples_per_second": 422.911,
+      "eval_steps_per_second": 13.216,
+      "step": 124500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.8269739682063025e-05,
+      "loss": 1.6585,
+      "step": 125000
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 1.465975046157837,
+      "eval_runtime": 236.3659,
+      "eval_samples_per_second": 423.073,
+      "eval_steps_per_second": 13.221,
+      "step": 125000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.822281864079128e-05,
+      "loss": 1.6569,
+      "step": 125500
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 1.4764187335968018,
+      "eval_runtime": 224.8017,
+      "eval_samples_per_second": 444.837,
+      "eval_steps_per_second": 13.901,
+      "step": 125500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.8175897599519525e-05,
+      "loss": 1.6557,
+      "step": 126000
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 1.4717729091644287,
+      "eval_runtime": 221.9623,
+      "eval_samples_per_second": 450.527,
+      "eval_steps_per_second": 14.079,
+      "step": 126000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.812897655824778e-05,
+      "loss": 1.6409,
+      "step": 126500
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.4730979204177856,
+      "eval_runtime": 221.9654,
+      "eval_samples_per_second": 450.521,
+      "eval_steps_per_second": 14.079,
+      "step": 126500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.808205551697603e-05,
+      "loss": 1.6487,
+      "step": 127000
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.4701308012008667,
+      "eval_runtime": 222.0375,
+      "eval_samples_per_second": 450.374,
+      "eval_steps_per_second": 14.074,
+      "step": 127000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.8035134475704285e-05,
+      "loss": 1.645,
+      "step": 127500
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.4616634845733643,
+      "eval_runtime": 222.0725,
+      "eval_samples_per_second": 450.303,
+      "eval_steps_per_second": 14.072,
+      "step": 127500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.798821343443254e-05,
+      "loss": 1.6402,
+      "step": 128000
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.4659932851791382,
+      "eval_runtime": 222.1612,
+      "eval_samples_per_second": 450.124,
+      "eval_steps_per_second": 14.066,
+      "step": 128000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.794129239316079e-05,
+      "loss": 1.6373,
+      "step": 128500
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.4619441032409668,
+      "eval_runtime": 222.2305,
+      "eval_samples_per_second": 449.983,
+      "eval_steps_per_second": 14.062,
+      "step": 128500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.789437135188904e-05,
+      "loss": 1.6473,
+      "step": 129000
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.4622013568878174,
+      "eval_runtime": 222.2288,
+      "eval_samples_per_second": 449.987,
+      "eval_steps_per_second": 14.062,
+      "step": 129000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.784745031061729e-05,
+      "loss": 1.6421,
+      "step": 129500
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.4600987434387207,
+      "eval_runtime": 222.2845,
+      "eval_samples_per_second": 449.874,
+      "eval_steps_per_second": 14.059,
+      "step": 129500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.7800529269345545e-05,
+      "loss": 1.6401,
+      "step": 130000
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.4546868801116943,
+      "eval_runtime": 229.4754,
+      "eval_samples_per_second": 435.777,
+      "eval_steps_per_second": 13.618,
+      "step": 130000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.77536082280738e-05,
+      "loss": 1.6446,
+      "step": 130500
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.4492428302764893,
+      "eval_runtime": 228.4821,
+      "eval_samples_per_second": 437.671,
+      "eval_steps_per_second": 13.677,
+      "step": 130500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.770668718680205e-05,
+      "loss": 1.6274,
+      "step": 131000
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.4530912637710571,
+      "eval_runtime": 228.7476,
+      "eval_samples_per_second": 437.163,
+      "eval_steps_per_second": 13.661,
+      "step": 131000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.76597661455303e-05,
+      "loss": 1.6548,
+      "step": 131500
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.453102946281433,
+      "eval_runtime": 232.9785,
+      "eval_samples_per_second": 429.224,
+      "eval_steps_per_second": 13.413,
+      "step": 131500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.761284510425855e-05,
+      "loss": 1.6128,
+      "step": 132000
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.445427656173706,
+      "eval_runtime": 232.8681,
+      "eval_samples_per_second": 429.428,
+      "eval_steps_per_second": 13.42,
+      "step": 132000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.7565924062986804e-05,
+      "loss": 1.6217,
+      "step": 132500
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.4480894804000854,
+      "eval_runtime": 229.5409,
+      "eval_samples_per_second": 435.652,
+      "eval_steps_per_second": 13.614,
+      "step": 132500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.751900302171506e-05,
+      "loss": 1.6292,
+      "step": 133000
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.4402817487716675,
+      "eval_runtime": 223.3255,
+      "eval_samples_per_second": 447.777,
+      "eval_steps_per_second": 13.993,
+      "step": 133000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.747208198044331e-05,
+      "loss": 1.6262,
+      "step": 133500
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.4454220533370972,
+      "eval_runtime": 223.1317,
+      "eval_samples_per_second": 448.166,
+      "eval_steps_per_second": 14.005,
+      "step": 133500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.7425160939171564e-05,
+      "loss": 1.6355,
+      "step": 134000
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.4490951299667358,
+      "eval_runtime": 222.8039,
+      "eval_samples_per_second": 448.825,
+      "eval_steps_per_second": 14.026,
+      "step": 134000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.737823989789981e-05,
+      "loss": 1.6361,
+      "step": 134500
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.446262001991272,
+      "eval_runtime": 222.7683,
+      "eval_samples_per_second": 448.897,
+      "eval_steps_per_second": 14.028,
+      "step": 134500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.7331318856628064e-05,
+      "loss": 1.6147,
+      "step": 135000
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.4396119117736816,
+      "eval_runtime": 222.7955,
+      "eval_samples_per_second": 448.842,
+      "eval_steps_per_second": 14.026,
+      "step": 135000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.728439781535632e-05,
+      "loss": 1.6442,
+      "step": 135500
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 1.4377689361572266,
+      "eval_runtime": 222.8075,
+      "eval_samples_per_second": 448.818,
+      "eval_steps_per_second": 14.026,
+      "step": 135500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.723747677408457e-05,
+      "loss": 1.6069,
+      "step": 136000
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.436558723449707,
+      "eval_runtime": 222.7834,
+      "eval_samples_per_second": 448.867,
+      "eval_steps_per_second": 14.027,
+      "step": 136000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.7190555732812824e-05,
+      "loss": 1.6183,
+      "step": 136500
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.4427673816680908,
+      "eval_runtime": 222.8527,
+      "eval_samples_per_second": 448.727,
+      "eval_steps_per_second": 14.023,
+      "step": 136500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.714363469154107e-05,
+      "loss": 1.6183,
+      "step": 137000
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.4434157609939575,
+      "eval_runtime": 222.9556,
+      "eval_samples_per_second": 448.52,
+      "eval_steps_per_second": 14.016,
+      "step": 137000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.7096713650269324e-05,
+      "loss": 1.6068,
+      "step": 137500
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.4364444017410278,
+      "eval_runtime": 222.8985,
+      "eval_samples_per_second": 448.635,
+      "eval_steps_per_second": 14.02,
+      "step": 137500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.704979260899758e-05,
+      "loss": 1.6363,
+      "step": 138000
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.4355006217956543,
+      "eval_runtime": 222.8369,
+      "eval_samples_per_second": 448.759,
+      "eval_steps_per_second": 14.024,
+      "step": 138000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.700287156772583e-05,
+      "loss": 1.6119,
+      "step": 138500
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.4310338497161865,
+      "eval_runtime": 222.7256,
+      "eval_samples_per_second": 448.983,
+      "eval_steps_per_second": 14.031,
+      "step": 138500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.6955950526454084e-05,
+      "loss": 1.6483,
+      "step": 139000
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.4362512826919556,
+      "eval_runtime": 222.6329,
+      "eval_samples_per_second": 449.17,
+      "eval_steps_per_second": 14.037,
+      "step": 139000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.690902948518234e-05,
+      "loss": 1.612,
+      "step": 139500
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 1.4332929849624634,
+      "eval_runtime": 222.5859,
+      "eval_samples_per_second": 449.265,
+      "eval_steps_per_second": 14.04,
+      "step": 139500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 3.686210844391059e-05,
+      "loss": 1.61,
+      "step": 140000
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 1.4201843738555908,
+      "eval_runtime": 226.6701,
+      "eval_samples_per_second": 441.17,
+      "eval_steps_per_second": 13.787,
+      "step": 140000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.681518740263884e-05,
+      "loss": 1.6236,
+      "step": 140500
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.4394793510437012,
+      "eval_runtime": 234.5172,
+      "eval_samples_per_second": 426.408,
+      "eval_steps_per_second": 13.325,
+      "step": 140500
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 3.676826636136709e-05,
+      "loss": 1.6185,
+      "step": 141000
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.433275818824768,
+      "eval_runtime": 234.8559,
+      "eval_samples_per_second": 425.793,
+      "eval_steps_per_second": 13.306,
+      "step": 141000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.672134532009534e-05,
+      "loss": 1.6061,
+      "step": 141500
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.418621301651001,
+      "eval_runtime": 236.7519,
+      "eval_samples_per_second": 422.383,
+      "eval_steps_per_second": 13.199,
+      "step": 141500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.6674424278823596e-05,
+      "loss": 1.6104,
+      "step": 142000
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.4307106733322144,
+      "eval_runtime": 235.4095,
+      "eval_samples_per_second": 424.792,
+      "eval_steps_per_second": 13.275,
+      "step": 142000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.662750323755185e-05,
+      "loss": 1.6079,
+      "step": 142500
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.4250682592391968,
+      "eval_runtime": 235.4454,
+      "eval_samples_per_second": 424.727,
+      "eval_steps_per_second": 13.273,
+      "step": 142500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 3.65805821962801e-05,
+      "loss": 1.5905,
+      "step": 143000
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.4250717163085938,
+      "eval_runtime": 235.7357,
+      "eval_samples_per_second": 424.204,
+      "eval_steps_per_second": 13.256,
+      "step": 143000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.653366115500835e-05,
+      "loss": 1.6081,
+      "step": 143500
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.4208674430847168,
+      "eval_runtime": 234.1473,
+      "eval_samples_per_second": 427.082,
+      "eval_steps_per_second": 13.346,
+      "step": 143500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.64867401137366e-05,
+      "loss": 1.6312,
+      "step": 144000
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.4182296991348267,
+      "eval_runtime": 235.783,
+      "eval_samples_per_second": 424.119,
+      "eval_steps_per_second": 13.254,
+      "step": 144000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.6439819072464856e-05,
+      "loss": 1.6118,
+      "step": 144500
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 1.4259034395217896,
+      "eval_runtime": 235.3478,
+      "eval_samples_per_second": 424.903,
+      "eval_steps_per_second": 13.278,
+      "step": 144500
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 3.639289803119311e-05,
+      "loss": 1.6056,
+      "step": 145000
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 1.4210768938064575,
+      "eval_runtime": 233.3721,
+      "eval_samples_per_second": 428.5,
+      "eval_steps_per_second": 13.391,
+      "step": 145000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.634597698992136e-05,
+      "loss": 1.6058,
+      "step": 145500
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.4050568342208862,
+      "eval_runtime": 231.2948,
+      "eval_samples_per_second": 432.349,
+      "eval_steps_per_second": 13.511,
+      "step": 145500
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.6299055948649616e-05,
+      "loss": 1.6005,
+      "step": 146000
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.4116266965866089,
+      "eval_runtime": 231.0158,
+      "eval_samples_per_second": 432.871,
+      "eval_steps_per_second": 13.527,
+      "step": 146000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 3.625213490737787e-05,
+      "loss": 1.5909,
+      "step": 146500
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 1.4132776260375977,
+      "eval_runtime": 231.0824,
+      "eval_samples_per_second": 432.746,
+      "eval_steps_per_second": 13.523,
+      "step": 146500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.6205213866106116e-05,
+      "loss": 1.614,
+      "step": 147000
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.413307547569275,
+      "eval_runtime": 222.5174,
+      "eval_samples_per_second": 449.403,
+      "eval_steps_per_second": 14.044,
+      "step": 147000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.615829282483437e-05,
+      "loss": 1.5907,
+      "step": 147500
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.4117494821548462,
+      "eval_runtime": 222.5274,
+      "eval_samples_per_second": 449.383,
+      "eval_steps_per_second": 14.043,
+      "step": 147500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.611137178356262e-05,
+      "loss": 1.5882,
+      "step": 148000
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.4053359031677246,
+      "eval_runtime": 222.505,
+      "eval_samples_per_second": 449.428,
+      "eval_steps_per_second": 14.045,
+      "step": 148000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 3.6064450742290876e-05,
+      "loss": 1.5864,
+      "step": 148500
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.3983163833618164,
+      "eval_runtime": 222.439,
+      "eval_samples_per_second": 449.561,
+      "eval_steps_per_second": 14.049,
+      "step": 148500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.601752970101913e-05,
+      "loss": 1.5881,
+      "step": 149000
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 1.4097754955291748,
+      "eval_runtime": 222.6731,
+      "eval_samples_per_second": 449.089,
+      "eval_steps_per_second": 14.034,
+      "step": 149000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 3.597060865974738e-05,
+      "loss": 1.5866,
+      "step": 149500
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 1.4093042612075806,
+      "eval_runtime": 222.7781,
+      "eval_samples_per_second": 448.877,
+      "eval_steps_per_second": 14.027,
+      "step": 149500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.5923687618475635e-05,
+      "loss": 1.5835,
+      "step": 150000
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.4054986238479614,
+      "eval_runtime": 228.6714,
+      "eval_samples_per_second": 437.309,
+      "eval_steps_per_second": 13.666,
+      "step": 150000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.587676657720388e-05,
+      "loss": 1.5792,
+      "step": 150500
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 1.40665864944458,
+      "eval_runtime": 228.4732,
+      "eval_samples_per_second": 437.688,
+      "eval_steps_per_second": 13.678,
+      "step": 150500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.5829845535932135e-05,
+      "loss": 1.5667,
+      "step": 151000
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.4050201177597046,
+      "eval_runtime": 228.2289,
+      "eval_samples_per_second": 438.157,
+      "eval_steps_per_second": 13.692,
+      "step": 151000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 3.578292449466039e-05,
+      "loss": 1.5771,
+      "step": 151500
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.4022421836853027,
+      "eval_runtime": 233.937,
+      "eval_samples_per_second": 427.466,
+      "eval_steps_per_second": 13.358,
+      "step": 151500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.573600345338864e-05,
+      "loss": 1.5768,
+      "step": 152000
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 1.4057670831680298,
+      "eval_runtime": 233.6648,
+      "eval_samples_per_second": 427.963,
+      "eval_steps_per_second": 13.374,
+      "step": 152000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 3.5689082412116895e-05,
+      "loss": 1.5752,
+      "step": 152500
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 1.403328537940979,
+      "eval_runtime": 239.3336,
+      "eval_samples_per_second": 417.827,
+      "eval_steps_per_second": 13.057,
+      "step": 152500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.564216137084515e-05,
+      "loss": 1.5757,
+      "step": 153000
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.3900455236434937,
+      "eval_runtime": 237.9128,
+      "eval_samples_per_second": 420.322,
+      "eval_steps_per_second": 13.135,
+      "step": 153000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.5595240329573395e-05,
+      "loss": 1.5689,
+      "step": 153500
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.3995317220687866,
+      "eval_runtime": 239.4513,
+      "eval_samples_per_second": 417.621,
+      "eval_steps_per_second": 13.051,
+      "step": 153500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.554831928830165e-05,
+      "loss": 1.5896,
+      "step": 154000
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.3904423713684082,
+      "eval_runtime": 236.1241,
+      "eval_samples_per_second": 423.506,
+      "eval_steps_per_second": 13.235,
+      "step": 154000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.55013982470299e-05,
+      "loss": 1.5916,
+      "step": 154500
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.400153636932373,
+      "eval_runtime": 227.755,
+      "eval_samples_per_second": 439.068,
+      "eval_steps_per_second": 13.721,
+      "step": 154500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 3.5454477205758155e-05,
+      "loss": 1.5896,
+      "step": 155000
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.3952091932296753,
+      "eval_runtime": 222.2884,
+      "eval_samples_per_second": 449.866,
+      "eval_steps_per_second": 14.058,
+      "step": 155000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.540755616448641e-05,
+      "loss": 1.5947,
+      "step": 155500
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.3903526067733765,
+      "eval_runtime": 222.5604,
+      "eval_samples_per_second": 449.316,
+      "eval_steps_per_second": 14.041,
+      "step": 155500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 3.536063512321466e-05,
+      "loss": 1.5658,
+      "step": 156000
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.388243317604065,
+      "eval_runtime": 222.4336,
+      "eval_samples_per_second": 449.572,
+      "eval_steps_per_second": 14.049,
+      "step": 156000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.531371408194291e-05,
+      "loss": 1.5711,
+      "step": 156500
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.3848968744277954,
+      "eval_runtime": 222.2993,
+      "eval_samples_per_second": 449.844,
+      "eval_steps_per_second": 14.058,
+      "step": 156500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.526679304067116e-05,
+      "loss": 1.5877,
+      "step": 157000
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.3827555179595947,
+      "eval_runtime": 222.1315,
+      "eval_samples_per_second": 450.184,
+      "eval_steps_per_second": 14.068,
+      "step": 157000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.5219871999399414e-05,
+      "loss": 1.5695,
+      "step": 157500
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.390235185623169,
+      "eval_runtime": 222.0358,
+      "eval_samples_per_second": 450.378,
+      "eval_steps_per_second": 14.074,
+      "step": 157500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 3.517295095812767e-05,
+      "loss": 1.5652,
+      "step": 158000
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.3825620412826538,
+      "eval_runtime": 221.9295,
+      "eval_samples_per_second": 450.594,
+      "eval_steps_per_second": 14.081,
+      "step": 158000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.512602991685592e-05,
+      "loss": 1.5696,
+      "step": 158500
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.3775691986083984,
+      "eval_runtime": 221.9726,
+      "eval_samples_per_second": 450.506,
+      "eval_steps_per_second": 14.078,
+      "step": 158500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.5079108875584174e-05,
+      "loss": 1.5738,
+      "step": 159000
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.3838963508605957,
+      "eval_runtime": 221.9255,
+      "eval_samples_per_second": 450.602,
+      "eval_steps_per_second": 14.081,
+      "step": 159000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.503218783431242e-05,
+      "loss": 1.5685,
+      "step": 159500
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.384406566619873,
+      "eval_runtime": 221.8809,
+      "eval_samples_per_second": 450.692,
+      "eval_steps_per_second": 14.084,
+      "step": 159500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 3.4985266793040674e-05,
+      "loss": 1.5653,
+      "step": 160000
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.3828558921813965,
+      "eval_runtime": 221.9019,
+      "eval_samples_per_second": 450.65,
+      "eval_steps_per_second": 14.083,
+      "step": 160000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.493834575176893e-05,
+      "loss": 1.5865,
+      "step": 160500
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 1.385338306427002,
+      "eval_runtime": 221.7161,
+      "eval_samples_per_second": 451.027,
+      "eval_steps_per_second": 14.095,
+      "step": 160500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 3.489142471049718e-05,
+      "loss": 1.5501,
+      "step": 161000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 1.3802553415298462,
+      "eval_runtime": 221.7084,
+      "eval_samples_per_second": 451.043,
+      "eval_steps_per_second": 14.095,
+      "step": 161000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.4844503669225434e-05,
+      "loss": 1.5556,
+      "step": 161500
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.3760793209075928,
+      "eval_runtime": 221.6941,
+      "eval_samples_per_second": 451.072,
+      "eval_steps_per_second": 14.096,
+      "step": 161500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.479758262795368e-05,
+      "loss": 1.5857,
+      "step": 162000
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.38301420211792,
+      "eval_runtime": 221.6934,
+      "eval_samples_per_second": 451.073,
+      "eval_steps_per_second": 14.096,
+      "step": 162000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.4750661586681934e-05,
+      "loss": 1.5549,
+      "step": 162500
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.3752098083496094,
+      "eval_runtime": 221.8522,
+      "eval_samples_per_second": 450.751,
+      "eval_steps_per_second": 14.086,
+      "step": 162500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.470374054541019e-05,
+      "loss": 1.5498,
+      "step": 163000
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 1.378463625907898,
+      "eval_runtime": 221.8169,
+      "eval_samples_per_second": 450.822,
+      "eval_steps_per_second": 14.088,
+      "step": 163000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 3.465681950413844e-05,
+      "loss": 1.5719,
+      "step": 163500
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 1.3748183250427246,
+      "eval_runtime": 221.783,
+      "eval_samples_per_second": 450.891,
+      "eval_steps_per_second": 14.09,
+      "step": 163500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.4609898462866694e-05,
+      "loss": 1.5751,
+      "step": 164000
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.3731558322906494,
+      "eval_runtime": 221.7643,
+      "eval_samples_per_second": 450.929,
+      "eval_steps_per_second": 14.092,
+      "step": 164000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 3.456297742159495e-05,
+      "loss": 1.5459,
+      "step": 164500
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.3722800016403198,
+      "eval_runtime": 221.6317,
+      "eval_samples_per_second": 451.199,
+      "eval_steps_per_second": 14.1,
+      "step": 164500
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.451605638032319e-05,
+      "loss": 1.5603,
+      "step": 165000
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.366198182106018,
+      "eval_runtime": 221.6334,
+      "eval_samples_per_second": 451.196,
+      "eval_steps_per_second": 14.1,
+      "step": 165000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.446913533905145e-05,
+      "loss": 1.5479,
+      "step": 165500
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.3646401166915894,
+      "eval_runtime": 221.6561,
+      "eval_samples_per_second": 451.149,
+      "eval_steps_per_second": 14.098,
+      "step": 165500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.44222142977797e-05,
+      "loss": 1.5386,
+      "step": 166000
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 1.3601208925247192,
+      "eval_runtime": 221.6415,
+      "eval_samples_per_second": 451.179,
+      "eval_steps_per_second": 14.099,
+      "step": 166000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 3.437529325650795e-05,
+      "loss": 1.5442,
+      "step": 166500
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 1.357480764389038,
+      "eval_runtime": 221.6464,
+      "eval_samples_per_second": 451.169,
+      "eval_steps_per_second": 14.099,
+      "step": 166500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.4328372215236207e-05,
+      "loss": 1.5476,
+      "step": 167000
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.3579912185668945,
+      "eval_runtime": 221.645,
+      "eval_samples_per_second": 451.172,
+      "eval_steps_per_second": 14.099,
+      "step": 167000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.428145117396445e-05,
+      "loss": 1.538,
+      "step": 167500
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.3654874563217163,
+      "eval_runtime": 221.6529,
+      "eval_samples_per_second": 451.156,
+      "eval_steps_per_second": 14.099,
+      "step": 167500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.4234530132692706e-05,
+      "loss": 1.5531,
+      "step": 168000
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.3746790885925293,
+      "eval_runtime": 221.6568,
+      "eval_samples_per_second": 451.148,
+      "eval_steps_per_second": 14.098,
+      "step": 168000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.418760909142096e-05,
+      "loss": 1.5538,
+      "step": 168500
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.3619598150253296,
+      "eval_runtime": 221.6175,
+      "eval_samples_per_second": 451.228,
+      "eval_steps_per_second": 14.101,
+      "step": 168500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.414068805014921e-05,
+      "loss": 1.5497,
+      "step": 169000
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.3550140857696533,
+      "eval_runtime": 221.6445,
+      "eval_samples_per_second": 451.173,
+      "eval_steps_per_second": 14.099,
+      "step": 169000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.4093767008877466e-05,
+      "loss": 1.5373,
+      "step": 169500
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.3558791875839233,
+      "eval_runtime": 221.8117,
+      "eval_samples_per_second": 450.833,
+      "eval_steps_per_second": 14.089,
+      "step": 169500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.404684596760572e-05,
+      "loss": 1.551,
+      "step": 170000
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.3610763549804688,
+      "eval_runtime": 221.8143,
+      "eval_samples_per_second": 450.828,
+      "eval_steps_per_second": 14.088,
+      "step": 170000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.3999924926333966e-05,
+      "loss": 1.5348,
+      "step": 170500
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.3658798933029175,
+      "eval_runtime": 221.7237,
+      "eval_samples_per_second": 451.012,
+      "eval_steps_per_second": 14.094,
+      "step": 170500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.395300388506222e-05,
+      "loss": 1.5391,
+      "step": 171000
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.3553804159164429,
+      "eval_runtime": 221.7256,
+      "eval_samples_per_second": 451.008,
+      "eval_steps_per_second": 14.094,
+      "step": 171000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.390608284379047e-05,
+      "loss": 1.5268,
+      "step": 171500
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 1.3583581447601318,
+      "eval_runtime": 221.5416,
+      "eval_samples_per_second": 451.383,
+      "eval_steps_per_second": 14.106,
+      "step": 171500
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.3859161802518726e-05,
+      "loss": 1.5377,
+      "step": 172000
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 1.3604706525802612,
+      "eval_runtime": 221.5467,
+      "eval_samples_per_second": 451.372,
+      "eval_steps_per_second": 14.105,
+      "step": 172000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.381224076124698e-05,
+      "loss": 1.5201,
+      "step": 172500
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.3594732284545898,
+      "eval_runtime": 221.5767,
+      "eval_samples_per_second": 451.311,
+      "eval_steps_per_second": 14.103,
+      "step": 172500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.3765319719975226e-05,
+      "loss": 1.5475,
+      "step": 173000
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.3577581644058228,
+      "eval_runtime": 221.5721,
+      "eval_samples_per_second": 451.32,
+      "eval_steps_per_second": 14.104,
+      "step": 173000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.371839867870348e-05,
+      "loss": 1.5442,
+      "step": 173500
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.3556467294692993,
+      "eval_runtime": 221.5702,
+      "eval_samples_per_second": 451.324,
+      "eval_steps_per_second": 14.104,
+      "step": 173500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.367147763743173e-05,
+      "loss": 1.5487,
+      "step": 174000
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.35114324092865,
+      "eval_runtime": 221.5518,
+      "eval_samples_per_second": 451.362,
+      "eval_steps_per_second": 14.105,
+      "step": 174000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.3624556596159985e-05,
+      "loss": 1.5282,
+      "step": 174500
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.3505486249923706,
+      "eval_runtime": 221.5706,
+      "eval_samples_per_second": 451.323,
+      "eval_steps_per_second": 14.104,
+      "step": 174500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.357763555488824e-05,
+      "loss": 1.5335,
+      "step": 175000
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.3498034477233887,
+      "eval_runtime": 221.5638,
+      "eval_samples_per_second": 451.337,
+      "eval_steps_per_second": 14.104,
+      "step": 175000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.353071451361649e-05,
+      "loss": 1.5217,
+      "step": 175500
+    },
+    {
+      "epoch": 1.65,
+      "eval_loss": 1.3372061252593994,
+      "eval_runtime": 221.6104,
+      "eval_samples_per_second": 451.242,
+      "eval_steps_per_second": 14.101,
+      "step": 175500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.348379347234474e-05,
+      "loss": 1.5374,
+      "step": 176000
+    },
+    {
+      "epoch": 1.65,
+      "eval_loss": 1.3543497323989868,
+      "eval_runtime": 221.5921,
+      "eval_samples_per_second": 451.28,
+      "eval_steps_per_second": 14.102,
+      "step": 176000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.343687243107299e-05,
+      "loss": 1.5282,
+      "step": 176500
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 1.3469191789627075,
+      "eval_runtime": 221.5417,
+      "eval_samples_per_second": 451.382,
+      "eval_steps_per_second": 14.106,
+      "step": 176500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.3389951389801245e-05,
+      "loss": 1.5135,
+      "step": 177000
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 1.3495582342147827,
+      "eval_runtime": 221.7135,
+      "eval_samples_per_second": 451.033,
+      "eval_steps_per_second": 14.095,
+      "step": 177000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.33430303485295e-05,
+      "loss": 1.5604,
+      "step": 177500
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 1.3418365716934204,
+      "eval_runtime": 221.6933,
+      "eval_samples_per_second": 451.074,
+      "eval_steps_per_second": 14.096,
+      "step": 177500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.329610930725775e-05,
+      "loss": 1.5222,
+      "step": 178000
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 1.3411223888397217,
+      "eval_runtime": 221.6176,
+      "eval_samples_per_second": 451.228,
+      "eval_steps_per_second": 14.101,
+      "step": 178000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3249188265986e-05,
+      "loss": 1.5207,
+      "step": 178500
+    },
+    {
+      "epoch": 1.68,
+      "eval_loss": 1.3367658853530884,
+      "eval_runtime": 221.5787,
+      "eval_samples_per_second": 451.307,
+      "eval_steps_per_second": 14.103,
+      "step": 178500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.320226722471425e-05,
+      "loss": 1.5232,
+      "step": 179000
+    },
+    {
+      "epoch": 1.68,
+      "eval_loss": 1.3388681411743164,
+      "eval_runtime": 221.3977,
+      "eval_samples_per_second": 451.676,
+      "eval_steps_per_second": 14.115,
+      "step": 179000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.3155346183442505e-05,
+      "loss": 1.5397,
+      "step": 179500
+    },
+    {
+      "epoch": 1.68,
+      "eval_loss": 1.3412151336669922,
+      "eval_runtime": 221.3909,
+      "eval_samples_per_second": 451.69,
+      "eval_steps_per_second": 14.115,
+      "step": 179500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.310842514217076e-05,
+      "loss": 1.5368,
+      "step": 180000
+    },
+    {
+      "epoch": 1.69,
+      "eval_loss": 1.3428925275802612,
+      "eval_runtime": 221.3914,
+      "eval_samples_per_second": 451.689,
+      "eval_steps_per_second": 14.115,
+      "step": 180000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.306150410089901e-05,
+      "loss": 1.5595,
+      "step": 180500
+    },
+    {
+      "epoch": 1.69,
+      "eval_loss": 1.336493730545044,
+      "eval_runtime": 221.2471,
+      "eval_samples_per_second": 451.983,
+      "eval_steps_per_second": 14.124,
+      "step": 180500
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.3014583059627265e-05,
+      "loss": 1.5407,
+      "step": 181000
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 1.3366632461547852,
+      "eval_runtime": 221.2531,
+      "eval_samples_per_second": 451.971,
+      "eval_steps_per_second": 14.124,
+      "step": 181000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.296766201835551e-05,
+      "loss": 1.5289,
+      "step": 181500
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 1.3385401964187622,
+      "eval_runtime": 221.3851,
+      "eval_samples_per_second": 451.702,
+      "eval_steps_per_second": 14.116,
+      "step": 181500
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.2920740977083764e-05,
+      "loss": 1.5259,
+      "step": 182000
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 1.3364156484603882,
+      "eval_runtime": 221.3895,
+      "eval_samples_per_second": 451.693,
+      "eval_steps_per_second": 14.115,
+      "step": 182000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.287381993581202e-05,
+      "loss": 1.5351,
+      "step": 182500
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 1.3378472328186035,
+      "eval_runtime": 221.3747,
+      "eval_samples_per_second": 451.723,
+      "eval_steps_per_second": 14.116,
+      "step": 182500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.282689889454027e-05,
+      "loss": 1.5167,
+      "step": 183000
+    },
+    {
+      "epoch": 1.72,
+      "eval_loss": 1.3326358795166016,
+      "eval_runtime": 221.3651,
+      "eval_samples_per_second": 451.742,
+      "eval_steps_per_second": 14.117,
+      "step": 183000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.2779977853268524e-05,
+      "loss": 1.5312,
+      "step": 183500
+    },
+    {
+      "epoch": 1.72,
+      "eval_loss": 1.3356856107711792,
+      "eval_runtime": 221.3292,
+      "eval_samples_per_second": 451.816,
+      "eval_steps_per_second": 14.119,
+      "step": 183500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.273305681199677e-05,
+      "loss": 1.5343,
+      "step": 184000
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 1.332999348640442,
+      "eval_runtime": 221.3046,
+      "eval_samples_per_second": 451.866,
+      "eval_steps_per_second": 14.121,
+      "step": 184000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.2686135770725024e-05,
+      "loss": 1.532,
+      "step": 184500
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 1.3290119171142578,
+      "eval_runtime": 221.4415,
+      "eval_samples_per_second": 451.586,
+      "eval_steps_per_second": 14.112,
+      "step": 184500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.263921472945328e-05,
+      "loss": 1.5106,
+      "step": 185000
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 1.3379125595092773,
+      "eval_runtime": 221.5598,
+      "eval_samples_per_second": 451.345,
+      "eval_steps_per_second": 14.105,
+      "step": 185000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.259229368818153e-05,
+      "loss": 1.5146,
+      "step": 185500
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 1.3202601671218872,
+      "eval_runtime": 221.5182,
+      "eval_samples_per_second": 451.43,
+      "eval_steps_per_second": 14.107,
+      "step": 185500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.2545372646909784e-05,
+      "loss": 1.5059,
+      "step": 186000
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 1.3281402587890625,
+      "eval_runtime": 221.4506,
+      "eval_samples_per_second": 451.568,
+      "eval_steps_per_second": 14.111,
+      "step": 186000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.249845160563804e-05,
+      "loss": 1.5206,
+      "step": 186500
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 1.3232953548431396,
+      "eval_runtime": 221.1927,
+      "eval_samples_per_second": 452.095,
+      "eval_steps_per_second": 14.128,
+      "step": 186500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.2451530564366284e-05,
+      "loss": 1.5332,
+      "step": 187000
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 1.328766107559204,
+      "eval_runtime": 221.1463,
+      "eval_samples_per_second": 452.189,
+      "eval_steps_per_second": 14.131,
+      "step": 187000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.240460952309454e-05,
+      "loss": 1.5253,
+      "step": 187500
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 1.3251526355743408,
+      "eval_runtime": 221.188,
+      "eval_samples_per_second": 452.104,
+      "eval_steps_per_second": 14.128,
+      "step": 187500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.235768848182279e-05,
+      "loss": 1.5412,
+      "step": 188000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 1.3271182775497437,
+      "eval_runtime": 221.1601,
+      "eval_samples_per_second": 452.161,
+      "eval_steps_per_second": 14.13,
+      "step": 188000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.2310767440551044e-05,
+      "loss": 1.5102,
+      "step": 188500
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 1.32902193069458,
+      "eval_runtime": 221.169,
+      "eval_samples_per_second": 452.143,
+      "eval_steps_per_second": 14.129,
+      "step": 188500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.22638463992793e-05,
+      "loss": 1.5158,
+      "step": 189000
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 1.3175742626190186,
+      "eval_runtime": 221.1728,
+      "eval_samples_per_second": 452.135,
+      "eval_steps_per_second": 14.129,
+      "step": 189000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.221692535800755e-05,
+      "loss": 1.5015,
+      "step": 189500
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 1.3242403268814087,
+      "eval_runtime": 221.1038,
+      "eval_samples_per_second": 452.276,
+      "eval_steps_per_second": 14.134,
+      "step": 189500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.21700043167358e-05,
+      "loss": 1.5226,
+      "step": 190000
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 1.3241665363311768,
+      "eval_runtime": 221.0819,
+      "eval_samples_per_second": 452.321,
+      "eval_steps_per_second": 14.135,
+      "step": 190000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.212308327546405e-05,
+      "loss": 1.5164,
+      "step": 190500
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.3255438804626465,
+      "eval_runtime": 221.0427,
+      "eval_samples_per_second": 452.401,
+      "eval_steps_per_second": 14.138,
+      "step": 190500
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.20761622341923e-05,
+      "loss": 1.4973,
+      "step": 191000
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.317428469657898,
+      "eval_runtime": 221.0182,
+      "eval_samples_per_second": 452.451,
+      "eval_steps_per_second": 14.139,
+      "step": 191000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.2029241192920556e-05,
+      "loss": 1.5136,
+      "step": 191500
+    },
+    {
+      "epoch": 1.8,
+      "eval_loss": 1.3230990171432495,
+      "eval_runtime": 221.0488,
+      "eval_samples_per_second": 452.389,
+      "eval_steps_per_second": 14.137,
+      "step": 191500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.198232015164881e-05,
+      "loss": 1.4964,
+      "step": 192000
+    },
+    {
+      "epoch": 1.8,
+      "eval_loss": 1.3161015510559082,
+      "eval_runtime": 221.0504,
+      "eval_samples_per_second": 452.385,
+      "eval_steps_per_second": 14.137,
+      "step": 192000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.1935399110377056e-05,
+      "loss": 1.5111,
+      "step": 192500
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 1.3233813047409058,
+      "eval_runtime": 221.2963,
+      "eval_samples_per_second": 451.883,
+      "eval_steps_per_second": 14.121,
+      "step": 192500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.188847806910531e-05,
+      "loss": 1.4848,
+      "step": 193000
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 1.3204160928726196,
+      "eval_runtime": 221.2711,
+      "eval_samples_per_second": 451.934,
+      "eval_steps_per_second": 14.123,
+      "step": 193000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.184155702783356e-05,
+      "loss": 1.5071,
+      "step": 193500
+    },
+    {
+      "epoch": 1.82,
+      "eval_loss": 1.3207734823226929,
+      "eval_runtime": 221.1959,
+      "eval_samples_per_second": 452.088,
+      "eval_steps_per_second": 14.128,
+      "step": 193500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.1794635986561816e-05,
+      "loss": 1.505,
+      "step": 194000
+    },
+    {
+      "epoch": 1.82,
+      "eval_loss": 1.3126252889633179,
+      "eval_runtime": 221.1627,
+      "eval_samples_per_second": 452.156,
+      "eval_steps_per_second": 14.13,
+      "step": 194000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.174771494529007e-05,
+      "loss": 1.5166,
+      "step": 194500
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 1.3139557838439941,
+      "eval_runtime": 221.0163,
+      "eval_samples_per_second": 452.455,
+      "eval_steps_per_second": 14.139,
+      "step": 194500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.170079390401832e-05,
+      "loss": 1.5,
+      "step": 195000
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 1.3125284910202026,
+      "eval_runtime": 221.0368,
+      "eval_samples_per_second": 452.413,
+      "eval_steps_per_second": 14.138,
+      "step": 195000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.165387286274657e-05,
+      "loss": 1.5055,
+      "step": 195500
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 1.3205658197402954,
+      "eval_runtime": 221.047,
+      "eval_samples_per_second": 452.392,
+      "eval_steps_per_second": 14.137,
+      "step": 195500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.160695182147482e-05,
+      "loss": 1.4874,
+      "step": 196000
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 1.3127025365829468,
+      "eval_runtime": 221.0478,
+      "eval_samples_per_second": 452.391,
+      "eval_steps_per_second": 14.137,
+      "step": 196000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.1560030780203076e-05,
+      "loss": 1.4879,
+      "step": 196500
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 1.3064494132995605,
+      "eval_runtime": 221.0498,
+      "eval_samples_per_second": 452.387,
+      "eval_steps_per_second": 14.137,
+      "step": 196500
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.151310973893133e-05,
+      "loss": 1.4988,
+      "step": 197000
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 1.3118481636047363,
+      "eval_runtime": 221.0593,
+      "eval_samples_per_second": 452.367,
+      "eval_steps_per_second": 14.136,
+      "step": 197000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.146618869765958e-05,
+      "loss": 1.5136,
+      "step": 197500
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 1.3082212209701538,
+      "eval_runtime": 221.0316,
+      "eval_samples_per_second": 452.424,
+      "eval_steps_per_second": 14.138,
+      "step": 197500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.141926765638783e-05,
+      "loss": 1.4772,
+      "step": 198000
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 1.3054986000061035,
+      "eval_runtime": 221.0487,
+      "eval_samples_per_second": 452.389,
+      "eval_steps_per_second": 14.137,
+      "step": 198000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.137234661511608e-05,
+      "loss": 1.4835,
+      "step": 198500
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 1.3038787841796875,
+      "eval_runtime": 221.0318,
+      "eval_samples_per_second": 452.424,
+      "eval_steps_per_second": 14.138,
+      "step": 198500
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.1325425573844335e-05,
+      "loss": 1.4953,
+      "step": 199000
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 1.3007208108901978,
+      "eval_runtime": 221.0384,
+      "eval_samples_per_second": 452.41,
+      "eval_steps_per_second": 14.138,
+      "step": 199000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.127850453257259e-05,
+      "loss": 1.488,
+      "step": 199500
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 1.301313042640686,
+      "eval_runtime": 221.0649,
+      "eval_samples_per_second": 452.356,
+      "eval_steps_per_second": 14.136,
+      "step": 199500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.123158349130084e-05,
+      "loss": 1.4808,
+      "step": 200000
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 1.307305097579956,
+      "eval_runtime": 221.0452,
+      "eval_samples_per_second": 452.396,
+      "eval_steps_per_second": 14.137,
+      "step": 200000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.1184662450029095e-05,
+      "loss": 1.49,
+      "step": 200500
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 1.3078263998031616,
+      "eval_runtime": 220.9228,
+      "eval_samples_per_second": 452.647,
+      "eval_steps_per_second": 14.145,
+      "step": 200500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.113774140875734e-05,
+      "loss": 1.4568,
+      "step": 201000
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 1.3042628765106201,
+      "eval_runtime": 221.0985,
+      "eval_samples_per_second": 452.287,
+      "eval_steps_per_second": 14.134,
+      "step": 201000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.1090820367485595e-05,
+      "loss": 1.4818,
+      "step": 201500
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 1.3147982358932495,
+      "eval_runtime": 221.0384,
+      "eval_samples_per_second": 452.41,
+      "eval_steps_per_second": 14.138,
+      "step": 201500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.104389932621385e-05,
+      "loss": 1.4693,
+      "step": 202000
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 1.2982258796691895,
+      "eval_runtime": 220.9646,
+      "eval_samples_per_second": 452.561,
+      "eval_steps_per_second": 14.143,
+      "step": 202000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.09969782849421e-05,
+      "loss": 1.4645,
+      "step": 202500
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 1.3004374504089355,
+      "eval_runtime": 220.9286,
+      "eval_samples_per_second": 452.635,
+      "eval_steps_per_second": 14.145,
+      "step": 202500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.0950057243670355e-05,
+      "loss": 1.4802,
+      "step": 203000
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 1.2999825477600098,
+      "eval_runtime": 220.8399,
+      "eval_samples_per_second": 452.817,
+      "eval_steps_per_second": 14.151,
+      "step": 203000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.09031362023986e-05,
+      "loss": 1.4797,
+      "step": 203500
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 1.2966198921203613,
+      "eval_runtime": 220.8339,
+      "eval_samples_per_second": 452.829,
+      "eval_steps_per_second": 14.151,
+      "step": 203500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.0856215161126855e-05,
+      "loss": 1.4995,
+      "step": 204000
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 1.2981046438217163,
+      "eval_runtime": 220.814,
+      "eval_samples_per_second": 452.87,
+      "eval_steps_per_second": 14.152,
+      "step": 204000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.080929411985511e-05,
+      "loss": 1.5087,
+      "step": 204500
+    },
+    {
+      "epoch": 1.92,
+      "eval_loss": 1.2951112985610962,
+      "eval_runtime": 220.8383,
+      "eval_samples_per_second": 452.82,
+      "eval_steps_per_second": 14.151,
+      "step": 204500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.076237307858336e-05,
+      "loss": 1.4984,
+      "step": 205000
+    },
+    {
+      "epoch": 1.92,
+      "eval_loss": 1.2964484691619873,
+      "eval_runtime": 220.8897,
+      "eval_samples_per_second": 452.715,
+      "eval_steps_per_second": 14.147,
+      "step": 205000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.0715452037311615e-05,
+      "loss": 1.4864,
+      "step": 205500
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.2925612926483154,
+      "eval_runtime": 220.8427,
+      "eval_samples_per_second": 452.811,
+      "eval_steps_per_second": 14.15,
+      "step": 205500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.066853099603987e-05,
+      "loss": 1.4752,
+      "step": 206000
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.3018134832382202,
+      "eval_runtime": 220.8565,
+      "eval_samples_per_second": 452.783,
+      "eval_steps_per_second": 14.149,
+      "step": 206000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.0621609954768114e-05,
+      "loss": 1.5018,
+      "step": 206500
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 1.2949315309524536,
+      "eval_runtime": 220.8631,
+      "eval_samples_per_second": 452.769,
+      "eval_steps_per_second": 14.149,
+      "step": 206500
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.057468891349637e-05,
+      "loss": 1.4902,
+      "step": 207000
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 1.2878872156143188,
+      "eval_runtime": 220.8333,
+      "eval_samples_per_second": 452.83,
+      "eval_steps_per_second": 14.151,
+      "step": 207000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.052776787222462e-05,
+      "loss": 1.4699,
+      "step": 207500
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 1.2887647151947021,
+      "eval_runtime": 220.8064,
+      "eval_samples_per_second": 452.885,
+      "eval_steps_per_second": 14.153,
+      "step": 207500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.0480846830952874e-05,
+      "loss": 1.4834,
+      "step": 208000
+    },
+    {
+      "epoch": 1.95,
+      "eval_loss": 1.290585994720459,
+      "eval_runtime": 220.7934,
+      "eval_samples_per_second": 452.912,
+      "eval_steps_per_second": 14.154,
+      "step": 208000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.0433925789681128e-05,
+      "loss": 1.4561,
+      "step": 208500
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 1.297606110572815,
+      "eval_runtime": 220.8078,
+      "eval_samples_per_second": 452.882,
+      "eval_steps_per_second": 14.153,
+      "step": 208500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.0387004748409374e-05,
+      "loss": 1.4673,
+      "step": 209000
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 1.2891823053359985,
+      "eval_runtime": 220.802,
+      "eval_samples_per_second": 452.895,
+      "eval_steps_per_second": 14.153,
+      "step": 209000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.0340083707137627e-05,
+      "loss": 1.4872,
+      "step": 209500
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 1.2896424531936646,
+      "eval_runtime": 220.9571,
+      "eval_samples_per_second": 452.577,
+      "eval_steps_per_second": 14.143,
+      "step": 209500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.029316266586588e-05,
+      "loss": 1.4538,
+      "step": 210000
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 1.2892200946807861,
+      "eval_runtime": 220.9468,
+      "eval_samples_per_second": 452.598,
+      "eval_steps_per_second": 14.144,
+      "step": 210000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0246241624594134e-05,
+      "loss": 1.4672,
+      "step": 210500
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 1.2804533243179321,
+      "eval_runtime": 220.8886,
+      "eval_samples_per_second": 452.717,
+      "eval_steps_per_second": 14.147,
+      "step": 210500
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0199320583322387e-05,
+      "loss": 1.46,
+      "step": 211000
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 1.292427659034729,
+      "eval_runtime": 220.8568,
+      "eval_samples_per_second": 452.782,
+      "eval_steps_per_second": 14.149,
+      "step": 211000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.015239954205064e-05,
+      "loss": 1.4597,
+      "step": 211500
+    },
+    {
+      "epoch": 1.98,
+      "eval_loss": 1.2845008373260498,
+      "eval_runtime": 220.714,
+      "eval_samples_per_second": 453.075,
+      "eval_steps_per_second": 14.159,
+      "step": 211500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0105478500778887e-05,
+      "loss": 1.4603,
+      "step": 212000
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.2904783487319946,
+      "eval_runtime": 220.7365,
+      "eval_samples_per_second": 453.029,
+      "eval_steps_per_second": 14.157,
+      "step": 212000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.005855745950714e-05,
+      "loss": 1.4812,
+      "step": 212500
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.2882287502288818,
+      "eval_runtime": 220.7381,
+      "eval_samples_per_second": 453.025,
+      "eval_steps_per_second": 14.157,
+      "step": 212500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0011636418235394e-05,
+      "loss": 1.4662,
+      "step": 213000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.277458667755127,
+      "eval_runtime": 220.7153,
+      "eval_samples_per_second": 453.072,
+      "eval_steps_per_second": 14.159,
+      "step": 213000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.9964715376963647e-05,
+      "loss": 1.4622,
+      "step": 213500
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.287771463394165,
+      "eval_runtime": 220.7913,
+      "eval_samples_per_second": 452.916,
+      "eval_steps_per_second": 14.154,
+      "step": 213500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.99177943356919e-05,
+      "loss": 1.4618,
+      "step": 214000
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 1.2776265144348145,
+      "eval_runtime": 220.7076,
+      "eval_samples_per_second": 453.088,
+      "eval_steps_per_second": 14.159,
+      "step": 214000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.987087329442015e-05,
+      "loss": 1.4462,
+      "step": 214500
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 1.282547116279602,
+      "eval_runtime": 220.7335,
+      "eval_samples_per_second": 453.035,
+      "eval_steps_per_second": 14.157,
+      "step": 214500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9823952253148403e-05,
+      "loss": 1.4502,
+      "step": 215000
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 1.2854583263397217,
+      "eval_runtime": 220.783,
+      "eval_samples_per_second": 452.933,
+      "eval_steps_per_second": 14.154,
+      "step": 215000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9777031211876653e-05,
+      "loss": 1.4419,
+      "step": 215500
+    },
+    {
+      "epoch": 2.02,
+      "eval_loss": 1.282270908355713,
+      "eval_runtime": 220.7936,
+      "eval_samples_per_second": 452.912,
+      "eval_steps_per_second": 14.153,
+      "step": 215500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9730110170604906e-05,
+      "loss": 1.4476,
+      "step": 216000
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 1.273934006690979,
+      "eval_runtime": 220.778,
+      "eval_samples_per_second": 452.944,
+      "eval_steps_per_second": 14.154,
+      "step": 216000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.968318912933316e-05,
+      "loss": 1.4549,
+      "step": 216500
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 1.2787511348724365,
+      "eval_runtime": 220.7612,
+      "eval_samples_per_second": 452.978,
+      "eval_steps_per_second": 14.156,
+      "step": 216500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9636268088061413e-05,
+      "loss": 1.4614,
+      "step": 217000
+    },
+    {
+      "epoch": 2.04,
+      "eval_loss": 1.2720485925674438,
+      "eval_runtime": 220.7698,
+      "eval_samples_per_second": 452.96,
+      "eval_steps_per_second": 14.155,
+      "step": 217000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9589347046789663e-05,
+      "loss": 1.4594,
+      "step": 217500
+    },
+    {
+      "epoch": 2.04,
+      "eval_loss": 1.2739953994750977,
+      "eval_runtime": 220.7583,
+      "eval_samples_per_second": 452.984,
+      "eval_steps_per_second": 14.156,
+      "step": 217500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.9542426005517916e-05,
+      "loss": 1.4633,
+      "step": 218000
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 1.279288411140442,
+      "eval_runtime": 220.7177,
+      "eval_samples_per_second": 453.068,
+      "eval_steps_per_second": 14.158,
+      "step": 218000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.949550496424617e-05,
+      "loss": 1.4755,
+      "step": 218500
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 1.2718831300735474,
+      "eval_runtime": 220.8318,
+      "eval_samples_per_second": 452.833,
+      "eval_steps_per_second": 14.151,
+      "step": 218500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.944858392297442e-05,
+      "loss": 1.4608,
+      "step": 219000
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 1.2677054405212402,
+      "eval_runtime": 220.8761,
+      "eval_samples_per_second": 452.742,
+      "eval_steps_per_second": 14.148,
+      "step": 219000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9401662881702673e-05,
+      "loss": 1.4327,
+      "step": 219500
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 1.270250678062439,
+      "eval_runtime": 220.795,
+      "eval_samples_per_second": 452.909,
+      "eval_steps_per_second": 14.153,
+      "step": 219500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9354741840430923e-05,
+      "loss": 1.4414,
+      "step": 220000
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 1.2740583419799805,
+      "eval_runtime": 220.7867,
+      "eval_samples_per_second": 452.926,
+      "eval_steps_per_second": 14.154,
+      "step": 220000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.9307820799159176e-05,
+      "loss": 1.4333,
+      "step": 220500
+    },
+    {
+      "epoch": 2.07,
+      "eval_loss": 1.2706983089447021,
+      "eval_runtime": 220.6245,
+      "eval_samples_per_second": 453.259,
+      "eval_steps_per_second": 14.164,
+      "step": 220500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.926089975788743e-05,
+      "loss": 1.445,
+      "step": 221000
+    },
+    {
+      "epoch": 2.07,
+      "eval_loss": 1.2743273973464966,
+      "eval_runtime": 220.6137,
+      "eval_samples_per_second": 453.281,
+      "eval_steps_per_second": 14.165,
+      "step": 221000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.9213978716615682e-05,
+      "loss": 1.4417,
+      "step": 221500
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 1.2643128633499146,
+      "eval_runtime": 220.619,
+      "eval_samples_per_second": 453.27,
+      "eval_steps_per_second": 14.165,
+      "step": 221500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.9167057675343932e-05,
+      "loss": 1.4456,
+      "step": 222000
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 1.2675349712371826,
+      "eval_runtime": 220.5897,
+      "eval_samples_per_second": 453.33,
+      "eval_steps_per_second": 14.167,
+      "step": 222000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.9120136634072186e-05,
+      "loss": 1.4515,
+      "step": 222500
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 1.2693637609481812,
+      "eval_runtime": 220.5837,
+      "eval_samples_per_second": 453.343,
+      "eval_steps_per_second": 14.167,
+      "step": 222500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.9073215592800436e-05,
+      "loss": 1.4649,
+      "step": 223000
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 1.269477367401123,
+      "eval_runtime": 220.5619,
+      "eval_samples_per_second": 453.387,
+      "eval_steps_per_second": 14.168,
+      "step": 223000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.902629455152869e-05,
+      "loss": 1.458,
+      "step": 223500
+    },
+    {
+      "epoch": 2.1,
+      "eval_loss": 1.270194411277771,
+      "eval_runtime": 220.6045,
+      "eval_samples_per_second": 453.3,
+      "eval_steps_per_second": 14.166,
+      "step": 223500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.8979373510256942e-05,
+      "loss": 1.4286,
+      "step": 224000
+    },
+    {
+      "epoch": 2.1,
+      "eval_loss": 1.2670238018035889,
+      "eval_runtime": 220.6056,
+      "eval_samples_per_second": 453.298,
+      "eval_steps_per_second": 14.166,
+      "step": 224000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.8932452468985195e-05,
+      "loss": 1.444,
+      "step": 224500
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 1.263913869857788,
+      "eval_runtime": 220.6041,
+      "eval_samples_per_second": 453.301,
+      "eval_steps_per_second": 14.166,
+      "step": 224500
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.888553142771345e-05,
+      "loss": 1.4352,
+      "step": 225000
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 1.2642849683761597,
+      "eval_runtime": 220.5676,
+      "eval_samples_per_second": 453.376,
+      "eval_steps_per_second": 14.168,
+      "step": 225000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.8838610386441695e-05,
+      "loss": 1.4535,
+      "step": 225500
+    },
+    {
+      "epoch": 2.12,
+      "eval_loss": 1.2708691358566284,
+      "eval_runtime": 220.6074,
+      "eval_samples_per_second": 453.294,
+      "eval_steps_per_second": 14.165,
+      "step": 225500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.879168934516995e-05,
+      "loss": 1.4289,
+      "step": 226000
+    },
+    {
+      "epoch": 2.12,
+      "eval_loss": 1.2640670537948608,
+      "eval_runtime": 220.5867,
+      "eval_samples_per_second": 453.336,
+      "eval_steps_per_second": 14.167,
+      "step": 226000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.8744768303898202e-05,
+      "loss": 1.4569,
+      "step": 226500
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 1.2591434717178345,
+      "eval_runtime": 220.5876,
+      "eval_samples_per_second": 453.335,
+      "eval_steps_per_second": 14.167,
+      "step": 226500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.8697847262626455e-05,
+      "loss": 1.4474,
+      "step": 227000
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 1.2669652700424194,
+      "eval_runtime": 220.5728,
+      "eval_samples_per_second": 453.365,
+      "eval_steps_per_second": 14.168,
+      "step": 227000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.865092622135471e-05,
+      "loss": 1.4412,
+      "step": 227500
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 1.2617195844650269,
+      "eval_runtime": 220.5569,
+      "eval_samples_per_second": 453.398,
+      "eval_steps_per_second": 14.169,
+      "step": 227500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.860400518008296e-05,
+      "loss": 1.4522,
+      "step": 228000
+    },
+    {
+      "epoch": 2.14,
+      "eval_loss": 1.262453317642212,
+      "eval_runtime": 220.6524,
+      "eval_samples_per_second": 453.201,
+      "eval_steps_per_second": 14.163,
+      "step": 228000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.8557084138811208e-05,
+      "loss": 1.4223,
+      "step": 228500
+    },
+    {
+      "epoch": 2.14,
+      "eval_loss": 1.2601869106292725,
+      "eval_runtime": 220.7247,
+      "eval_samples_per_second": 453.053,
+      "eval_steps_per_second": 14.158,
+      "step": 228500
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.851016309753946e-05,
+      "loss": 1.4352,
+      "step": 229000
+    },
+    {
+      "epoch": 2.15,
+      "eval_loss": 1.2503169775009155,
+      "eval_runtime": 220.6544,
+      "eval_samples_per_second": 453.197,
+      "eval_steps_per_second": 14.162,
+      "step": 229000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.8463242056267715e-05,
+      "loss": 1.4153,
+      "step": 229500
+    },
+    {
+      "epoch": 2.15,
+      "eval_loss": 1.2586244344711304,
+      "eval_runtime": 220.6119,
+      "eval_samples_per_second": 453.285,
+      "eval_steps_per_second": 14.165,
+      "step": 229500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.8416321014995968e-05,
+      "loss": 1.4255,
+      "step": 230000
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 1.2628742456436157,
+      "eval_runtime": 220.5148,
+      "eval_samples_per_second": 453.484,
+      "eval_steps_per_second": 14.171,
+      "step": 230000
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.836939997372422e-05,
+      "loss": 1.4402,
+      "step": 230500
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 1.2623021602630615,
+      "eval_runtime": 220.3405,
+      "eval_samples_per_second": 453.843,
+      "eval_steps_per_second": 14.183,
+      "step": 230500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.8322478932452475e-05,
+      "loss": 1.4327,
+      "step": 231000
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 1.2573188543319702,
+      "eval_runtime": 220.4056,
+      "eval_samples_per_second": 453.709,
+      "eval_steps_per_second": 14.178,
+      "step": 231000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.827555789118072e-05,
+      "loss": 1.4522,
+      "step": 231500
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 1.262677550315857,
+      "eval_runtime": 220.4232,
+      "eval_samples_per_second": 453.673,
+      "eval_steps_per_second": 14.177,
+      "step": 231500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.8228636849908974e-05,
+      "loss": 1.4528,
+      "step": 232000
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 1.259798526763916,
+      "eval_runtime": 220.4521,
+      "eval_samples_per_second": 453.613,
+      "eval_steps_per_second": 14.175,
+      "step": 232000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.8181715808637228e-05,
+      "loss": 1.4322,
+      "step": 232500
+    },
+    {
+      "epoch": 2.18,
+      "eval_loss": 1.2561190128326416,
+      "eval_runtime": 220.4464,
+      "eval_samples_per_second": 453.625,
+      "eval_steps_per_second": 14.176,
+      "step": 232500
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.813479476736548e-05,
+      "loss": 1.4285,
+      "step": 233000
+    },
+    {
+      "epoch": 2.19,
+      "eval_loss": 1.262929916381836,
+      "eval_runtime": 220.4749,
+      "eval_samples_per_second": 453.566,
+      "eval_steps_per_second": 14.174,
+      "step": 233000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.8087873726093734e-05,
+      "loss": 1.4357,
+      "step": 233500
+    },
+    {
+      "epoch": 2.19,
+      "eval_loss": 1.2526911497116089,
+      "eval_runtime": 220.5247,
+      "eval_samples_per_second": 453.464,
+      "eval_steps_per_second": 14.171,
+      "step": 233500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.804095268482198e-05,
+      "loss": 1.4242,
+      "step": 234000
+    },
+    {
+      "epoch": 2.2,
+      "eval_loss": 1.2577769756317139,
+      "eval_runtime": 220.5683,
+      "eval_samples_per_second": 453.374,
+      "eval_steps_per_second": 14.168,
+      "step": 234000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.7994031643550234e-05,
+      "loss": 1.4162,
+      "step": 234500
+    },
+    {
+      "epoch": 2.2,
+      "eval_loss": 1.2519463300704956,
+      "eval_runtime": 220.6043,
+      "eval_samples_per_second": 453.3,
+      "eval_steps_per_second": 14.166,
+      "step": 234500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.7947110602278487e-05,
+      "loss": 1.4332,
+      "step": 235000
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 1.2554024457931519,
+      "eval_runtime": 220.6422,
+      "eval_samples_per_second": 453.222,
+      "eval_steps_per_second": 14.163,
+      "step": 235000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.790018956100674e-05,
+      "loss": 1.4313,
+      "step": 235500
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 1.2516902685165405,
+      "eval_runtime": 220.661,
+      "eval_samples_per_second": 453.184,
+      "eval_steps_per_second": 14.162,
+      "step": 235500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.7853268519734994e-05,
+      "loss": 1.4295,
+      "step": 236000
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 1.246413230895996,
+      "eval_runtime": 220.6917,
+      "eval_samples_per_second": 453.121,
+      "eval_steps_per_second": 14.16,
+      "step": 236000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.7806347478463247e-05,
+      "loss": 1.4206,
+      "step": 236500
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 1.2490031719207764,
+      "eval_runtime": 220.6934,
+      "eval_samples_per_second": 453.117,
+      "eval_steps_per_second": 14.16,
+      "step": 236500
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.7759426437191494e-05,
+      "loss": 1.4304,
+      "step": 237000
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 1.2488088607788086,
+      "eval_runtime": 220.7139,
+      "eval_samples_per_second": 453.075,
+      "eval_steps_per_second": 14.159,
+      "step": 237000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.7712505395919747e-05,
+      "loss": 1.4205,
+      "step": 237500
+    },
+    {
+      "epoch": 2.23,
+      "eval_loss": 1.2458001375198364,
+      "eval_runtime": 220.8036,
+      "eval_samples_per_second": 452.891,
+      "eval_steps_per_second": 14.153,
+      "step": 237500
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.7665584354648e-05,
+      "loss": 1.4287,
+      "step": 238000
+    },
+    {
+      "epoch": 2.23,
+      "eval_loss": 1.2452707290649414,
+      "eval_runtime": 220.9497,
+      "eval_samples_per_second": 452.592,
+      "eval_steps_per_second": 14.143,
+      "step": 238000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.7618663313376253e-05,
+      "loss": 1.4384,
+      "step": 238500
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 1.2455310821533203,
+      "eval_runtime": 220.9115,
+      "eval_samples_per_second": 452.67,
+      "eval_steps_per_second": 14.146,
+      "step": 238500
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.7571742272104507e-05,
+      "loss": 1.4297,
+      "step": 239000
+    },
+    {
+      "epoch": 2.24,
+      "eval_loss": 1.2461364269256592,
+      "eval_runtime": 220.8712,
+      "eval_samples_per_second": 452.753,
+      "eval_steps_per_second": 14.149,
+      "step": 239000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.7524821230832753e-05,
+      "loss": 1.4192,
+      "step": 239500
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 1.241681456565857,
+      "eval_runtime": 220.7997,
+      "eval_samples_per_second": 452.899,
+      "eval_steps_per_second": 14.153,
+      "step": 239500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.7477900189561007e-05,
+      "loss": 1.4371,
+      "step": 240000
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 1.241499662399292,
+      "eval_runtime": 220.8006,
+      "eval_samples_per_second": 452.897,
+      "eval_steps_per_second": 14.153,
+      "step": 240000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.743097914828926e-05,
+      "loss": 1.4134,
+      "step": 240500
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 1.2505980730056763,
+      "eval_runtime": 220.9405,
+      "eval_samples_per_second": 452.611,
+      "eval_steps_per_second": 14.144,
+      "step": 240500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.7384058107017513e-05,
+      "loss": 1.4282,
+      "step": 241000
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 1.2383712530136108,
+      "eval_runtime": 220.9785,
+      "eval_samples_per_second": 452.533,
+      "eval_steps_per_second": 14.142,
+      "step": 241000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.7337137065745766e-05,
+      "loss": 1.423,
+      "step": 241500
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 1.2391642332077026,
+      "eval_runtime": 221.0121,
+      "eval_samples_per_second": 452.464,
+      "eval_steps_per_second": 14.14,
+      "step": 241500
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.729021602447402e-05,
+      "loss": 1.3965,
+      "step": 242000
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 1.234950304031372,
+      "eval_runtime": 221.0193,
+      "eval_samples_per_second": 452.449,
+      "eval_steps_per_second": 14.139,
+      "step": 242000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.7243294983202266e-05,
+      "loss": 1.4213,
+      "step": 242500
+    },
+    {
+      "epoch": 2.28,
+      "eval_loss": 1.2397407293319702,
+      "eval_runtime": 221.049,
+      "eval_samples_per_second": 452.388,
+      "eval_steps_per_second": 14.137,
+      "step": 242500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.719637394193052e-05,
+      "loss": 1.4205,
+      "step": 243000
+    },
+    {
+      "epoch": 2.28,
+      "eval_loss": 1.239810824394226,
+      "eval_runtime": 221.0771,
+      "eval_samples_per_second": 452.331,
+      "eval_steps_per_second": 14.135,
+      "step": 243000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7149452900658773e-05,
+      "loss": 1.4043,
+      "step": 243500
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 1.234268307685852,
+      "eval_runtime": 221.0918,
+      "eval_samples_per_second": 452.301,
+      "eval_steps_per_second": 14.134,
+      "step": 243500
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7102531859387026e-05,
+      "loss": 1.4253,
+      "step": 244000
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 1.2382601499557495,
+      "eval_runtime": 221.0904,
+      "eval_samples_per_second": 452.304,
+      "eval_steps_per_second": 14.134,
+      "step": 244000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.705561081811528e-05,
+      "loss": 1.417,
+      "step": 244500
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 1.2414758205413818,
+      "eval_runtime": 221.0991,
+      "eval_samples_per_second": 452.286,
+      "eval_steps_per_second": 14.134,
+      "step": 244500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7008689776843526e-05,
+      "loss": 1.4246,
+      "step": 245000
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 1.2350417375564575,
+      "eval_runtime": 221.1225,
+      "eval_samples_per_second": 452.238,
+      "eval_steps_per_second": 14.132,
+      "step": 245000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.696176873557178e-05,
+      "loss": 1.4098,
+      "step": 245500
+    },
+    {
+      "epoch": 2.3,
+      "eval_loss": 1.231085181236267,
+      "eval_runtime": 221.136,
+      "eval_samples_per_second": 452.21,
+      "eval_steps_per_second": 14.132,
+      "step": 245500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6914847694300032e-05,
+      "loss": 1.4127,
+      "step": 246000
+    },
+    {
+      "epoch": 2.31,
+      "eval_loss": 1.2361029386520386,
+      "eval_runtime": 221.7853,
+      "eval_samples_per_second": 450.887,
+      "eval_steps_per_second": 14.09,
+      "step": 246000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6867926653028286e-05,
+      "loss": 1.3934,
+      "step": 246500
+    },
+    {
+      "epoch": 2.31,
+      "eval_loss": 1.2260022163391113,
+      "eval_runtime": 221.7286,
+      "eval_samples_per_second": 451.002,
+      "eval_steps_per_second": 14.094,
+      "step": 246500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.682100561175654e-05,
+      "loss": 1.4177,
+      "step": 247000
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 1.2279936075210571,
+      "eval_runtime": 221.7401,
+      "eval_samples_per_second": 450.979,
+      "eval_steps_per_second": 14.093,
+      "step": 247000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.6774084570484792e-05,
+      "loss": 1.393,
+      "step": 247500
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 1.230800986289978,
+      "eval_runtime": 221.7152,
+      "eval_samples_per_second": 451.029,
+      "eval_steps_per_second": 14.095,
+      "step": 247500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.672716352921304e-05,
+      "loss": 1.4007,
+      "step": 248000
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 1.2362948656082153,
+      "eval_runtime": 221.5844,
+      "eval_samples_per_second": 451.295,
+      "eval_steps_per_second": 14.103,
+      "step": 248000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.6680242487941292e-05,
+      "loss": 1.4119,
+      "step": 248500
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 1.2321358919143677,
+      "eval_runtime": 221.5867,
+      "eval_samples_per_second": 451.291,
+      "eval_steps_per_second": 14.103,
+      "step": 248500
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.6633321446669545e-05,
+      "loss": 1.4134,
+      "step": 249000
+    },
+    {
+      "epoch": 2.34,
+      "eval_loss": 1.2312347888946533,
+      "eval_runtime": 221.6239,
+      "eval_samples_per_second": 451.215,
+      "eval_steps_per_second": 14.1,
+      "step": 249000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.65864004053978e-05,
+      "loss": 1.3995,
+      "step": 249500
+    },
+    {
+      "epoch": 2.34,
+      "eval_loss": 1.227805495262146,
+      "eval_runtime": 221.6235,
+      "eval_samples_per_second": 451.216,
+      "eval_steps_per_second": 14.1,
+      "step": 249500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.6539479364126052e-05,
+      "loss": 1.4042,
+      "step": 250000
+    },
+    {
+      "epoch": 2.35,
+      "eval_loss": 1.2257417440414429,
+      "eval_runtime": 221.6629,
+      "eval_samples_per_second": 451.136,
+      "eval_steps_per_second": 14.098,
+      "step": 250000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.64925583228543e-05,
+      "loss": 1.4041,
+      "step": 250500
+    },
+    {
+      "epoch": 2.35,
+      "eval_loss": 1.2193177938461304,
+      "eval_runtime": 221.6777,
+      "eval_samples_per_second": 451.105,
+      "eval_steps_per_second": 14.097,
+      "step": 250500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6445637281582552e-05,
+      "loss": 1.4014,
+      "step": 251000
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 1.2312824726104736,
+      "eval_runtime": 221.7704,
+      "eval_samples_per_second": 450.917,
+      "eval_steps_per_second": 14.091,
+      "step": 251000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6398716240310805e-05,
+      "loss": 1.4122,
+      "step": 251500
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 1.2306718826293945,
+      "eval_runtime": 221.7735,
+      "eval_samples_per_second": 450.911,
+      "eval_steps_per_second": 14.091,
+      "step": 251500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.6351795199039058e-05,
+      "loss": 1.3839,
+      "step": 252000
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 1.2296006679534912,
+      "eval_runtime": 221.8207,
+      "eval_samples_per_second": 450.815,
+      "eval_steps_per_second": 14.088,
+      "step": 252000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.630487415776731e-05,
+      "loss": 1.4113,
+      "step": 252500
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.2209049463272095,
+      "eval_runtime": 221.8254,
+      "eval_samples_per_second": 450.805,
+      "eval_steps_per_second": 14.088,
+      "step": 252500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.6257953116495565e-05,
+      "loss": 1.3818,
+      "step": 253000
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.2337490320205688,
+      "eval_runtime": 221.8579,
+      "eval_samples_per_second": 450.739,
+      "eval_steps_per_second": 14.086,
+      "step": 253000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.621103207522381e-05,
+      "loss": 1.4197,
+      "step": 253500
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 1.2286659479141235,
+      "eval_runtime": 222.042,
+      "eval_samples_per_second": 450.365,
+      "eval_steps_per_second": 14.074,
+      "step": 253500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.6164111033952065e-05,
+      "loss": 1.407,
+      "step": 254000
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 1.2235952615737915,
+      "eval_runtime": 222.0342,
+      "eval_samples_per_second": 450.381,
+      "eval_steps_per_second": 14.074,
+      "step": 254000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.6117189992680318e-05,
+      "loss": 1.4031,
+      "step": 254500
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 1.2171984910964966,
+      "eval_runtime": 222.0173,
+      "eval_samples_per_second": 450.415,
+      "eval_steps_per_second": 14.075,
+      "step": 254500
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.607026895140857e-05,
+      "loss": 1.39,
+      "step": 255000
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 1.2256048917770386,
+      "eval_runtime": 221.9334,
+      "eval_samples_per_second": 450.586,
+      "eval_steps_per_second": 14.081,
+      "step": 255000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6023347910136824e-05,
+      "loss": 1.3946,
+      "step": 255500
+    },
+    {
+      "epoch": 2.4,
+      "eval_loss": 1.2248618602752686,
+      "eval_runtime": 221.838,
+      "eval_samples_per_second": 450.779,
+      "eval_steps_per_second": 14.087,
+      "step": 255500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.597642686886507e-05,
+      "loss": 1.417,
+      "step": 256000
+    },
+    {
+      "epoch": 2.4,
+      "eval_loss": 1.2168443202972412,
+      "eval_runtime": 221.8642,
+      "eval_samples_per_second": 450.726,
+      "eval_steps_per_second": 14.085,
+      "step": 256000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.5929505827593324e-05,
+      "loss": 1.4065,
+      "step": 256500
+    },
+    {
+      "epoch": 2.41,
+      "eval_loss": 1.2248560190200806,
+      "eval_runtime": 221.8502,
+      "eval_samples_per_second": 450.755,
+      "eval_steps_per_second": 14.086,
+      "step": 256500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.5882584786321578e-05,
+      "loss": 1.4046,
+      "step": 257000
+    },
+    {
+      "epoch": 2.41,
+      "eval_loss": 1.2218291759490967,
+      "eval_runtime": 221.8436,
+      "eval_samples_per_second": 450.768,
+      "eval_steps_per_second": 14.086,
+      "step": 257000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.583566374504983e-05,
+      "loss": 1.4101,
+      "step": 257500
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 1.2202662229537964,
+      "eval_runtime": 221.8576,
+      "eval_samples_per_second": 450.74,
+      "eval_steps_per_second": 14.086,
+      "step": 257500
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.5788742703778084e-05,
+      "loss": 1.3997,
+      "step": 258000
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 1.2214444875717163,
+      "eval_runtime": 221.8565,
+      "eval_samples_per_second": 450.742,
+      "eval_steps_per_second": 14.086,
+      "step": 258000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.5741821662506337e-05,
+      "loss": 1.4134,
+      "step": 258500
+    },
+    {
+      "epoch": 2.43,
+      "eval_loss": 1.220430612564087,
+      "eval_runtime": 221.8826,
+      "eval_samples_per_second": 450.689,
+      "eval_steps_per_second": 14.084,
+      "step": 258500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.5694900621234584e-05,
+      "loss": 1.3823,
+      "step": 259000
+    },
+    {
+      "epoch": 2.43,
+      "eval_loss": 1.2182176113128662,
+      "eval_runtime": 221.9231,
+      "eval_samples_per_second": 450.606,
+      "eval_steps_per_second": 14.081,
+      "step": 259000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.5647979579962837e-05,
+      "loss": 1.3854,
+      "step": 259500
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 1.2200777530670166,
+      "eval_runtime": 221.9107,
+      "eval_samples_per_second": 450.632,
+      "eval_steps_per_second": 14.082,
+      "step": 259500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.560105853869109e-05,
+      "loss": 1.4126,
+      "step": 260000
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 1.2206027507781982,
+      "eval_runtime": 221.8721,
+      "eval_samples_per_second": 450.71,
+      "eval_steps_per_second": 14.085,
+      "step": 260000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.5554137497419344e-05,
+      "loss": 1.3986,
+      "step": 260500
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 1.2171374559402466,
+      "eval_runtime": 222.0524,
+      "eval_samples_per_second": 450.344,
+      "eval_steps_per_second": 14.073,
+      "step": 260500
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5507216456147597e-05,
+      "loss": 1.401,
+      "step": 261000
+    },
+    {
+      "epoch": 2.45,
+      "eval_loss": 1.220663070678711,
+      "eval_runtime": 222.0724,
+      "eval_samples_per_second": 450.304,
+      "eval_steps_per_second": 14.072,
+      "step": 261000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5460295414875847e-05,
+      "loss": 1.3983,
+      "step": 261500
+    },
+    {
+      "epoch": 2.45,
+      "eval_loss": 1.2131149768829346,
+      "eval_runtime": 222.0269,
+      "eval_samples_per_second": 450.396,
+      "eval_steps_per_second": 14.075,
+      "step": 261500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5413374373604097e-05,
+      "loss": 1.3837,
+      "step": 262000
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 1.2130707502365112,
+      "eval_runtime": 222.0086,
+      "eval_samples_per_second": 450.433,
+      "eval_steps_per_second": 14.076,
+      "step": 262000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.536645333233235e-05,
+      "loss": 1.3935,
+      "step": 262500
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 1.2063870429992676,
+      "eval_runtime": 221.8375,
+      "eval_samples_per_second": 450.78,
+      "eval_steps_per_second": 14.087,
+      "step": 262500
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.5319532291060603e-05,
+      "loss": 1.4004,
+      "step": 263000
+    },
+    {
+      "epoch": 2.47,
+      "eval_loss": 1.21162748336792,
+      "eval_runtime": 221.8742,
+      "eval_samples_per_second": 450.706,
+      "eval_steps_per_second": 14.085,
+      "step": 263000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.5272611249788857e-05,
+      "loss": 1.3934,
+      "step": 263500
+    },
+    {
+      "epoch": 2.47,
+      "eval_loss": 1.2112177610397339,
+      "eval_runtime": 221.8733,
+      "eval_samples_per_second": 450.708,
+      "eval_steps_per_second": 14.085,
+      "step": 263500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.522569020851711e-05,
+      "loss": 1.3747,
+      "step": 264000
+    },
+    {
+      "epoch": 2.48,
+      "eval_loss": 1.217407464981079,
+      "eval_runtime": 221.8721,
+      "eval_samples_per_second": 450.71,
+      "eval_steps_per_second": 14.085,
+      "step": 264000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.517876916724536e-05,
+      "loss": 1.3891,
+      "step": 264500
+    },
+    {
+      "epoch": 2.48,
+      "eval_loss": 1.215196132659912,
+      "eval_runtime": 221.9107,
+      "eval_samples_per_second": 450.632,
+      "eval_steps_per_second": 14.082,
+      "step": 264500
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.5131848125973613e-05,
+      "loss": 1.3753,
+      "step": 265000
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 1.2109936475753784,
+      "eval_runtime": 221.8926,
+      "eval_samples_per_second": 450.668,
+      "eval_steps_per_second": 14.083,
+      "step": 265000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.5084927084701863e-05,
+      "loss": 1.4029,
+      "step": 265500
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 1.2136719226837158,
+      "eval_runtime": 221.9088,
+      "eval_samples_per_second": 450.636,
+      "eval_steps_per_second": 14.082,
+      "step": 265500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.5038006043430116e-05,
+      "loss": 1.3716,
+      "step": 266000
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 1.2064990997314453,
+      "eval_runtime": 221.917,
+      "eval_samples_per_second": 450.619,
+      "eval_steps_per_second": 14.082,
+      "step": 266000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.499108500215837e-05,
+      "loss": 1.3917,
+      "step": 266500
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 1.2140556573867798,
+      "eval_runtime": 221.8926,
+      "eval_samples_per_second": 450.668,
+      "eval_steps_per_second": 14.083,
+      "step": 266500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.494416396088662e-05,
+      "loss": 1.3958,
+      "step": 267000
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 1.2115426063537598,
+      "eval_runtime": 221.881,
+      "eval_samples_per_second": 450.692,
+      "eval_steps_per_second": 14.084,
+      "step": 267000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4897242919614873e-05,
+      "loss": 1.388,
+      "step": 267500
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 1.2005212306976318,
+      "eval_runtime": 222.0705,
+      "eval_samples_per_second": 450.307,
+      "eval_steps_per_second": 14.072,
+      "step": 267500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.4850321878343126e-05,
+      "loss": 1.3769,
+      "step": 268000
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 1.2068332433700562,
+      "eval_runtime": 222.0339,
+      "eval_samples_per_second": 450.382,
+      "eval_steps_per_second": 14.074,
+      "step": 268000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.480340083707138e-05,
+      "loss": 1.4106,
+      "step": 268500
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 1.2093212604522705,
+      "eval_runtime": 222.0101,
+      "eval_samples_per_second": 450.43,
+      "eval_steps_per_second": 14.076,
+      "step": 268500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.475647979579963e-05,
+      "loss": 1.3883,
+      "step": 269000
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 1.2118886709213257,
+      "eval_runtime": 221.9501,
+      "eval_samples_per_second": 450.552,
+      "eval_steps_per_second": 14.08,
+      "step": 269000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4709558754527883e-05,
+      "loss": 1.3742,
+      "step": 269500
+    },
+    {
+      "epoch": 2.53,
+      "eval_loss": 1.2088725566864014,
+      "eval_runtime": 221.8528,
+      "eval_samples_per_second": 450.749,
+      "eval_steps_per_second": 14.086,
+      "step": 269500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.4662637713256136e-05,
+      "loss": 1.3922,
+      "step": 270000
+    },
+    {
+      "epoch": 2.53,
+      "eval_loss": 1.2059705257415771,
+      "eval_runtime": 221.9028,
+      "eval_samples_per_second": 450.648,
+      "eval_steps_per_second": 14.083,
+      "step": 270000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4615716671984386e-05,
+      "loss": 1.3924,
+      "step": 270500
+    },
+    {
+      "epoch": 2.54,
+      "eval_loss": 1.2069438695907593,
+      "eval_runtime": 221.8382,
+      "eval_samples_per_second": 450.779,
+      "eval_steps_per_second": 14.087,
+      "step": 270500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.456879563071264e-05,
+      "loss": 1.4043,
+      "step": 271000
+    },
+    {
+      "epoch": 2.54,
+      "eval_loss": 1.2041146755218506,
+      "eval_runtime": 221.8591,
+      "eval_samples_per_second": 450.736,
+      "eval_steps_per_second": 14.086,
+      "step": 271000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.4521874589440892e-05,
+      "loss": 1.3877,
+      "step": 271500
+    },
+    {
+      "epoch": 2.55,
+      "eval_loss": 1.209468960762024,
+      "eval_runtime": 221.837,
+      "eval_samples_per_second": 450.781,
+      "eval_steps_per_second": 14.087,
+      "step": 271500
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.4474953548169142e-05,
+      "loss": 1.3866,
+      "step": 272000
+    },
+    {
+      "epoch": 2.55,
+      "eval_loss": 1.2078853845596313,
+      "eval_runtime": 221.8543,
+      "eval_samples_per_second": 450.746,
+      "eval_steps_per_second": 14.086,
+      "step": 272000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.4428032506897396e-05,
+      "loss": 1.3892,
+      "step": 272500
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 1.203126311302185,
+      "eval_runtime": 221.8674,
+      "eval_samples_per_second": 450.72,
+      "eval_steps_per_second": 14.085,
+      "step": 272500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.438111146562565e-05,
+      "loss": 1.3876,
+      "step": 273000
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 1.208136796951294,
+      "eval_runtime": 221.8861,
+      "eval_samples_per_second": 450.682,
+      "eval_steps_per_second": 14.084,
+      "step": 273000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.43341904243539e-05,
+      "loss": 1.3791,
+      "step": 273500
+    },
+    {
+      "epoch": 2.57,
+      "eval_loss": 1.2000367641448975,
+      "eval_runtime": 221.8811,
+      "eval_samples_per_second": 450.692,
+      "eval_steps_per_second": 14.084,
+      "step": 273500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.4287269383082152e-05,
+      "loss": 1.3721,
+      "step": 274000
+    },
+    {
+      "epoch": 2.57,
+      "eval_loss": 1.1960883140563965,
+      "eval_runtime": 221.8902,
+      "eval_samples_per_second": 450.673,
+      "eval_steps_per_second": 14.084,
+      "step": 274000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.4240348341810405e-05,
+      "loss": 1.3758,
+      "step": 274500
+    },
+    {
+      "epoch": 2.58,
+      "eval_loss": 1.1972496509552002,
+      "eval_runtime": 222.0858,
+      "eval_samples_per_second": 450.276,
+      "eval_steps_per_second": 14.071,
+      "step": 274500
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.4193427300538655e-05,
+      "loss": 1.3854,
+      "step": 275000
+    },
+    {
+      "epoch": 2.58,
+      "eval_loss": 1.2011773586273193,
+      "eval_runtime": 222.0189,
+      "eval_samples_per_second": 450.412,
+      "eval_steps_per_second": 14.075,
+      "step": 275000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.414650625926691e-05,
+      "loss": 1.3815,
+      "step": 275500
+    },
+    {
+      "epoch": 2.59,
+      "eval_loss": 1.200337290763855,
+      "eval_runtime": 222.0085,
+      "eval_samples_per_second": 450.433,
+      "eval_steps_per_second": 14.076,
+      "step": 275500
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.409958521799516e-05,
+      "loss": 1.37,
+      "step": 276000
+    },
+    {
+      "epoch": 2.59,
+      "eval_loss": 1.203967809677124,
+      "eval_runtime": 221.9584,
+      "eval_samples_per_second": 450.535,
+      "eval_steps_per_second": 14.079,
+      "step": 276000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.405266417672341e-05,
+      "loss": 1.3913,
+      "step": 276500
+    },
+    {
+      "epoch": 2.59,
+      "eval_loss": 1.193058729171753,
+      "eval_runtime": 221.8004,
+      "eval_samples_per_second": 450.856,
+      "eval_steps_per_second": 14.089,
+      "step": 276500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.4005743135451665e-05,
+      "loss": 1.3707,
+      "step": 277000
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 1.2021199464797974,
+      "eval_runtime": 221.8184,
+      "eval_samples_per_second": 450.819,
+      "eval_steps_per_second": 14.088,
+      "step": 277000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.3958822094179915e-05,
+      "loss": 1.3903,
+      "step": 277500
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 1.1956475973129272,
+      "eval_runtime": 221.7913,
+      "eval_samples_per_second": 450.874,
+      "eval_steps_per_second": 14.09,
+      "step": 277500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.3911901052908168e-05,
+      "loss": 1.3792,
+      "step": 278000
+    },
+    {
+      "epoch": 2.61,
+      "eval_loss": 1.1976265907287598,
+      "eval_runtime": 221.8164,
+      "eval_samples_per_second": 450.823,
+      "eval_steps_per_second": 14.088,
+      "step": 278000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.386498001163642e-05,
+      "loss": 1.3765,
+      "step": 278500
+    },
+    {
+      "epoch": 2.61,
+      "eval_loss": 1.2006570100784302,
+      "eval_runtime": 221.8144,
+      "eval_samples_per_second": 450.827,
+      "eval_steps_per_second": 14.088,
+      "step": 278500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.381805897036467e-05,
+      "loss": 1.3589,
+      "step": 279000
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 1.1913626194000244,
+      "eval_runtime": 221.8059,
+      "eval_samples_per_second": 450.845,
+      "eval_steps_per_second": 14.089,
+      "step": 279000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.3771137929092925e-05,
+      "loss": 1.3887,
+      "step": 279500
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 1.1945520639419556,
+      "eval_runtime": 221.8425,
+      "eval_samples_per_second": 450.77,
+      "eval_steps_per_second": 14.087,
+      "step": 279500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.3724216887821178e-05,
+      "loss": 1.3897,
+      "step": 280000
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 1.1894328594207764,
+      "eval_runtime": 221.8345,
+      "eval_samples_per_second": 450.786,
+      "eval_steps_per_second": 14.087,
+      "step": 280000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.3677295846549428e-05,
+      "loss": 1.3875,
+      "step": 280500
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 1.190177083015442,
+      "eval_runtime": 221.8514,
+      "eval_samples_per_second": 450.752,
+      "eval_steps_per_second": 14.086,
+      "step": 280500
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.363037480527768e-05,
+      "loss": 1.3796,
+      "step": 281000
+    },
+    {
+      "epoch": 2.64,
+      "eval_loss": 1.1883279085159302,
+      "eval_runtime": 221.8103,
+      "eval_samples_per_second": 450.836,
+      "eval_steps_per_second": 14.089,
+      "step": 281000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.3583453764005934e-05,
+      "loss": 1.3563,
+      "step": 281500
+    },
+    {
+      "epoch": 2.64,
+      "eval_loss": 1.1856937408447266,
+      "eval_runtime": 222.0332,
+      "eval_samples_per_second": 450.383,
+      "eval_steps_per_second": 14.074,
+      "step": 281500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3536532722734184e-05,
+      "loss": 1.3546,
+      "step": 282000
+    },
+    {
+      "epoch": 2.65,
+      "eval_loss": 1.1944658756256104,
+      "eval_runtime": 221.9876,
+      "eval_samples_per_second": 450.476,
+      "eval_steps_per_second": 14.077,
+      "step": 282000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.3489611681462438e-05,
+      "loss": 1.3597,
+      "step": 282500
+    },
+    {
+      "epoch": 2.65,
+      "eval_loss": 1.1892412900924683,
+      "eval_runtime": 221.924,
+      "eval_samples_per_second": 450.605,
+      "eval_steps_per_second": 14.081,
+      "step": 282500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.3442690640190687e-05,
+      "loss": 1.368,
+      "step": 283000
+    },
+    {
+      "epoch": 2.66,
+      "eval_loss": 1.1899220943450928,
+      "eval_runtime": 221.8992,
+      "eval_samples_per_second": 450.655,
+      "eval_steps_per_second": 14.083,
+      "step": 283000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.339576959891894e-05,
+      "loss": 1.3766,
+      "step": 283500
+    },
+    {
+      "epoch": 2.66,
+      "eval_loss": 1.1951682567596436,
+      "eval_runtime": 221.7577,
+      "eval_samples_per_second": 450.943,
+      "eval_steps_per_second": 14.092,
+      "step": 283500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.3348848557647194e-05,
+      "loss": 1.369,
+      "step": 284000
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.1905887126922607,
+      "eval_runtime": 221.7249,
+      "eval_samples_per_second": 451.009,
+      "eval_steps_per_second": 14.094,
+      "step": 284000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.3301927516375444e-05,
+      "loss": 1.3634,
+      "step": 284500
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.1932637691497803,
+      "eval_runtime": 221.7372,
+      "eval_samples_per_second": 450.984,
+      "eval_steps_per_second": 14.093,
+      "step": 284500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.3255006475103697e-05,
+      "loss": 1.3571,
+      "step": 285000
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.1916605234146118,
+      "eval_runtime": 221.7465,
+      "eval_samples_per_second": 450.965,
+      "eval_steps_per_second": 14.093,
+      "step": 285000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.320808543383195e-05,
+      "loss": 1.3663,
+      "step": 285500
+    },
+    {
+      "epoch": 2.68,
+      "eval_loss": 1.1863957643508911,
+      "eval_runtime": 221.7586,
+      "eval_samples_per_second": 450.941,
+      "eval_steps_per_second": 14.092,
+      "step": 285500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.31611643925602e-05,
+      "loss": 1.3437,
+      "step": 286000
+    },
+    {
+      "epoch": 2.68,
+      "eval_loss": 1.186560034751892,
+      "eval_runtime": 221.7879,
+      "eval_samples_per_second": 450.881,
+      "eval_steps_per_second": 14.09,
+      "step": 286000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.3114243351288454e-05,
+      "loss": 1.3614,
+      "step": 286500
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 1.1834124326705933,
+      "eval_runtime": 221.7759,
+      "eval_samples_per_second": 450.906,
+      "eval_steps_per_second": 14.091,
+      "step": 286500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.3067322310016707e-05,
+      "loss": 1.3561,
+      "step": 287000
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 1.1831755638122559,
+      "eval_runtime": 221.7594,
+      "eval_samples_per_second": 450.939,
+      "eval_steps_per_second": 14.092,
+      "step": 287000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.3020401268744957e-05,
+      "loss": 1.3634,
+      "step": 287500
+    },
+    {
+      "epoch": 2.7,
+      "eval_loss": 1.1881248950958252,
+      "eval_runtime": 221.7452,
+      "eval_samples_per_second": 450.968,
+      "eval_steps_per_second": 14.093,
+      "step": 287500
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.297348022747321e-05,
+      "loss": 1.3574,
+      "step": 288000
+    },
+    {
+      "epoch": 2.7,
+      "eval_loss": 1.1852173805236816,
+      "eval_runtime": 221.7114,
+      "eval_samples_per_second": 451.037,
+      "eval_steps_per_second": 14.095,
+      "step": 288000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.292655918620146e-05,
+      "loss": 1.377,
+      "step": 288500
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 1.1913198232650757,
+      "eval_runtime": 221.8189,
+      "eval_samples_per_second": 450.818,
+      "eval_steps_per_second": 14.088,
+      "step": 288500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.2879638144929713e-05,
+      "loss": 1.3511,
+      "step": 289000
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 1.1883938312530518,
+      "eval_runtime": 221.884,
+      "eval_samples_per_second": 450.686,
+      "eval_steps_per_second": 14.084,
+      "step": 289000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.2832717103657967e-05,
+      "loss": 1.3486,
+      "step": 289500
+    },
+    {
+      "epoch": 2.72,
+      "eval_loss": 1.185508370399475,
+      "eval_runtime": 221.7951,
+      "eval_samples_per_second": 450.867,
+      "eval_steps_per_second": 14.09,
+      "step": 289500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.2785796062386216e-05,
+      "loss": 1.3584,
+      "step": 290000
+    },
+    {
+      "epoch": 2.72,
+      "eval_loss": 1.1850634813308716,
+      "eval_runtime": 221.7253,
+      "eval_samples_per_second": 451.009,
+      "eval_steps_per_second": 14.094,
+      "step": 290000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.273887502111447e-05,
+      "loss": 1.3634,
+      "step": 290500
+    },
+    {
+      "epoch": 2.73,
+      "eval_loss": 1.1823251247406006,
+      "eval_runtime": 221.5435,
+      "eval_samples_per_second": 451.379,
+      "eval_steps_per_second": 14.106,
+      "step": 290500
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.2691953979842723e-05,
+      "loss": 1.3652,
+      "step": 291000
+    },
+    {
+      "epoch": 2.73,
+      "eval_loss": 1.181983232498169,
+      "eval_runtime": 221.5124,
+      "eval_samples_per_second": 451.442,
+      "eval_steps_per_second": 14.108,
+      "step": 291000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2645032938570973e-05,
+      "loss": 1.3612,
+      "step": 291500
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 1.182550311088562,
+      "eval_runtime": 221.4655,
+      "eval_samples_per_second": 451.538,
+      "eval_steps_per_second": 14.111,
+      "step": 291500
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.2598111897299226e-05,
+      "loss": 1.3827,
+      "step": 292000
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 1.1845837831497192,
+      "eval_runtime": 221.4927,
+      "eval_samples_per_second": 451.482,
+      "eval_steps_per_second": 14.109,
+      "step": 292000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.255119085602748e-05,
+      "loss": 1.3722,
+      "step": 292500
+    },
+    {
+      "epoch": 2.74,
+      "eval_loss": 1.1796345710754395,
+      "eval_runtime": 221.4954,
+      "eval_samples_per_second": 451.477,
+      "eval_steps_per_second": 14.109,
+      "step": 292500
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.250426981475573e-05,
+      "loss": 1.3608,
+      "step": 293000
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 1.1794308423995972,
+      "eval_runtime": 221.4654,
+      "eval_samples_per_second": 451.538,
+      "eval_steps_per_second": 14.111,
+      "step": 293000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.2457348773483983e-05,
+      "loss": 1.3601,
+      "step": 293500
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 1.1854252815246582,
+      "eval_runtime": 221.506,
+      "eval_samples_per_second": 451.455,
+      "eval_steps_per_second": 14.108,
+      "step": 293500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.2410427732212233e-05,
+      "loss": 1.3723,
+      "step": 294000
+    },
+    {
+      "epoch": 2.76,
+      "eval_loss": 1.1760536432266235,
+      "eval_runtime": 221.4862,
+      "eval_samples_per_second": 451.495,
+      "eval_steps_per_second": 14.109,
+      "step": 294000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.2363506690940486e-05,
+      "loss": 1.3559,
+      "step": 294500
+    },
+    {
+      "epoch": 2.76,
+      "eval_loss": 1.1737596988677979,
+      "eval_runtime": 221.496,
+      "eval_samples_per_second": 451.475,
+      "eval_steps_per_second": 14.109,
+      "step": 294500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.231658564966874e-05,
+      "loss": 1.3596,
+      "step": 295000
+    },
+    {
+      "epoch": 2.77,
+      "eval_loss": 1.1840267181396484,
+      "eval_runtime": 221.4671,
+      "eval_samples_per_second": 451.534,
+      "eval_steps_per_second": 14.11,
+      "step": 295000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.226966460839699e-05,
+      "loss": 1.3469,
+      "step": 295500
+    },
+    {
+      "epoch": 2.77,
+      "eval_loss": 1.1722980737686157,
+      "eval_runtime": 221.4589,
+      "eval_samples_per_second": 451.551,
+      "eval_steps_per_second": 14.111,
+      "step": 295500
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.2222743567125242e-05,
+      "loss": 1.3636,
+      "step": 296000
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 1.1743708848953247,
+      "eval_runtime": 221.6515,
+      "eval_samples_per_second": 451.159,
+      "eval_steps_per_second": 14.099,
+      "step": 296000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.2175822525853496e-05,
+      "loss": 1.3522,
+      "step": 296500
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 1.1773130893707275,
+      "eval_runtime": 221.623,
+      "eval_samples_per_second": 451.217,
+      "eval_steps_per_second": 14.101,
+      "step": 296500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.2128901484581746e-05,
+      "loss": 1.3459,
+      "step": 297000
+    },
+    {
+      "epoch": 2.79,
+      "eval_loss": 1.1674100160598755,
+      "eval_runtime": 221.5803,
+      "eval_samples_per_second": 451.304,
+      "eval_steps_per_second": 14.103,
+      "step": 297000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.208198044331e-05,
+      "loss": 1.3634,
+      "step": 297500
+    },
+    {
+      "epoch": 2.79,
+      "eval_loss": 1.1741117238998413,
+      "eval_runtime": 221.517,
+      "eval_samples_per_second": 451.433,
+      "eval_steps_per_second": 14.107,
+      "step": 297500
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2035059402038252e-05,
+      "loss": 1.3529,
+      "step": 298000
+    },
+    {
+      "epoch": 2.8,
+      "eval_loss": 1.1743310689926147,
+      "eval_runtime": 221.3651,
+      "eval_samples_per_second": 451.742,
+      "eval_steps_per_second": 14.117,
+      "step": 298000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1988138360766502e-05,
+      "loss": 1.3524,
+      "step": 298500
+    },
+    {
+      "epoch": 2.8,
+      "eval_loss": 1.172131896018982,
+      "eval_runtime": 221.343,
+      "eval_samples_per_second": 451.788,
+      "eval_steps_per_second": 14.118,
+      "step": 298500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1941217319494755e-05,
+      "loss": 1.349,
+      "step": 299000
+    },
+    {
+      "epoch": 2.81,
+      "eval_loss": 1.1744838953018188,
+      "eval_runtime": 221.3399,
+      "eval_samples_per_second": 451.794,
+      "eval_steps_per_second": 14.119,
+      "step": 299000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1894296278223005e-05,
+      "loss": 1.3658,
+      "step": 299500
+    },
+    {
+      "epoch": 2.81,
+      "eval_loss": 1.167363166809082,
+      "eval_runtime": 221.371,
+      "eval_samples_per_second": 451.73,
+      "eval_steps_per_second": 14.117,
+      "step": 299500
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.184737523695126e-05,
+      "loss": 1.3663,
+      "step": 300000
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.1738266944885254,
+      "eval_runtime": 221.3431,
+      "eval_samples_per_second": 451.787,
+      "eval_steps_per_second": 14.118,
+      "step": 300000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.1800454195679512e-05,
+      "loss": 1.3473,
+      "step": 300500
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.173740267753601,
+      "eval_runtime": 221.0442,
+      "eval_samples_per_second": 452.398,
+      "eval_steps_per_second": 14.137,
+      "step": 300500
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.175353315440776e-05,
+      "loss": 1.3589,
+      "step": 301000
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.1683049201965332,
+      "eval_runtime": 221.0696,
+      "eval_samples_per_second": 452.346,
+      "eval_steps_per_second": 14.136,
+      "step": 301000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1706612113136015e-05,
+      "loss": 1.347,
+      "step": 301500
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 1.1638891696929932,
+      "eval_runtime": 221.0645,
+      "eval_samples_per_second": 452.357,
+      "eval_steps_per_second": 14.136,
+      "step": 301500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.1659691071864268e-05,
+      "loss": 1.3421,
+      "step": 302000
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 1.1762233972549438,
+      "eval_runtime": 221.1041,
+      "eval_samples_per_second": 452.276,
+      "eval_steps_per_second": 14.134,
+      "step": 302000
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.1612770030592518e-05,
+      "loss": 1.3507,
+      "step": 302500
+    },
+    {
+      "epoch": 2.84,
+      "eval_loss": 1.17283034324646,
+      "eval_runtime": 221.1012,
+      "eval_samples_per_second": 452.282,
+      "eval_steps_per_second": 14.134,
+      "step": 302500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.156584898932077e-05,
+      "loss": 1.3247,
+      "step": 303000
+    },
+    {
+      "epoch": 2.84,
+      "eval_loss": 1.1722640991210938,
+      "eval_runtime": 221.0672,
+      "eval_samples_per_second": 452.351,
+      "eval_steps_per_second": 14.136,
+      "step": 303000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1518927948049025e-05,
+      "loss": 1.3373,
+      "step": 303500
+    },
+    {
+      "epoch": 2.85,
+      "eval_loss": 1.1675173044204712,
+      "eval_runtime": 221.061,
+      "eval_samples_per_second": 452.364,
+      "eval_steps_per_second": 14.136,
+      "step": 303500
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1472006906777275e-05,
+      "loss": 1.347,
+      "step": 304000
+    },
+    {
+      "epoch": 2.85,
+      "eval_loss": 1.169805645942688,
+      "eval_runtime": 221.2478,
+      "eval_samples_per_second": 451.982,
+      "eval_steps_per_second": 14.124,
+      "step": 304000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.1425085865505528e-05,
+      "loss": 1.3549,
+      "step": 304500
+    },
+    {
+      "epoch": 2.86,
+      "eval_loss": 1.1683489084243774,
+      "eval_runtime": 221.2569,
+      "eval_samples_per_second": 451.963,
+      "eval_steps_per_second": 14.124,
+      "step": 304500
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.137816482423378e-05,
+      "loss": 1.3406,
+      "step": 305000
+    },
+    {
+      "epoch": 2.86,
+      "eval_loss": 1.1658775806427002,
+      "eval_runtime": 221.1911,
+      "eval_samples_per_second": 452.098,
+      "eval_steps_per_second": 14.128,
+      "step": 305000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.133124378296203e-05,
+      "loss": 1.3497,
+      "step": 305500
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 1.1645755767822266,
+      "eval_runtime": 221.1253,
+      "eval_samples_per_second": 452.232,
+      "eval_steps_per_second": 14.132,
+      "step": 305500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.1284322741690284e-05,
+      "loss": 1.3615,
+      "step": 306000
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 1.1647412776947021,
+      "eval_runtime": 220.9938,
+      "eval_samples_per_second": 452.501,
+      "eval_steps_per_second": 14.141,
+      "step": 306000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.1237401700418534e-05,
+      "loss": 1.3281,
+      "step": 306500
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 1.1673572063446045,
+      "eval_runtime": 220.9974,
+      "eval_samples_per_second": 452.494,
+      "eval_steps_per_second": 14.14,
+      "step": 306500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.1190480659146787e-05,
+      "loss": 1.3447,
+      "step": 307000
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 1.1644020080566406,
+      "eval_runtime": 221.0059,
+      "eval_samples_per_second": 452.477,
+      "eval_steps_per_second": 14.14,
+      "step": 307000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.114355961787504e-05,
+      "loss": 1.3518,
+      "step": 307500
+    },
+    {
+      "epoch": 2.89,
+      "eval_loss": 1.1606210470199585,
+      "eval_runtime": 220.9981,
+      "eval_samples_per_second": 452.493,
+      "eval_steps_per_second": 14.14,
+      "step": 307500
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.109663857660329e-05,
+      "loss": 1.3268,
+      "step": 308000
+    },
+    {
+      "epoch": 2.89,
+      "eval_loss": 1.172353982925415,
+      "eval_runtime": 220.9928,
+      "eval_samples_per_second": 452.503,
+      "eval_steps_per_second": 14.141,
+      "step": 308000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1049717535331544e-05,
+      "loss": 1.3281,
+      "step": 308500
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 1.1636762619018555,
+      "eval_runtime": 221.0284,
+      "eval_samples_per_second": 452.431,
+      "eval_steps_per_second": 14.138,
+      "step": 308500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1002796494059797e-05,
+      "loss": 1.3361,
+      "step": 309000
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 1.1628670692443848,
+      "eval_runtime": 221.0125,
+      "eval_samples_per_second": 452.463,
+      "eval_steps_per_second": 14.139,
+      "step": 309000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.0955875452788047e-05,
+      "loss": 1.3297,
+      "step": 309500
+    },
+    {
+      "epoch": 2.9,
+      "eval_loss": 1.1588451862335205,
+      "eval_runtime": 221.003,
+      "eval_samples_per_second": 452.482,
+      "eval_steps_per_second": 14.14,
+      "step": 309500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.09089544115163e-05,
+      "loss": 1.3492,
+      "step": 310000
+    },
+    {
+      "epoch": 2.91,
+      "eval_loss": 1.1633425951004028,
+      "eval_runtime": 220.9707,
+      "eval_samples_per_second": 452.549,
+      "eval_steps_per_second": 14.142,
+      "step": 310000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.0862033370244554e-05,
+      "loss": 1.3419,
+      "step": 310500
+    },
+    {
+      "epoch": 2.91,
+      "eval_loss": 1.160988211631775,
+      "eval_runtime": 220.9533,
+      "eval_samples_per_second": 452.584,
+      "eval_steps_per_second": 14.143,
+      "step": 310500
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.0815112328972804e-05,
+      "loss": 1.3277,
+      "step": 311000
+    },
+    {
+      "epoch": 2.92,
+      "eval_loss": 1.1551365852355957,
+      "eval_runtime": 220.9421,
+      "eval_samples_per_second": 452.607,
+      "eval_steps_per_second": 14.144,
+      "step": 311000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.0768191287701057e-05,
+      "loss": 1.3402,
+      "step": 311500
+    },
+    {
+      "epoch": 2.92,
+      "eval_loss": 1.163130283355713,
+      "eval_runtime": 220.9627,
+      "eval_samples_per_second": 452.565,
+      "eval_steps_per_second": 14.143,
+      "step": 311500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.072127024642931e-05,
+      "loss": 1.329,
+      "step": 312000
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 1.1555798053741455,
+      "eval_runtime": 221.0607,
+      "eval_samples_per_second": 452.364,
+      "eval_steps_per_second": 14.136,
+      "step": 312000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.067434920515756e-05,
+      "loss": 1.3467,
+      "step": 312500
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 1.1523679494857788,
+      "eval_runtime": 221.1292,
+      "eval_samples_per_second": 452.224,
+      "eval_steps_per_second": 14.132,
+      "step": 312500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0627428163885813e-05,
+      "loss": 1.3574,
+      "step": 313000
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 1.159830093383789,
+      "eval_runtime": 221.0551,
+      "eval_samples_per_second": 452.376,
+      "eval_steps_per_second": 14.137,
+      "step": 313000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0580507122614067e-05,
+      "loss": 1.343,
+      "step": 313500
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 1.1504855155944824,
+      "eval_runtime": 221.0384,
+      "eval_samples_per_second": 452.41,
+      "eval_steps_per_second": 14.138,
+      "step": 313500
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.0533586081342317e-05,
+      "loss": 1.3234,
+      "step": 314000
+    },
+    {
+      "epoch": 2.95,
+      "eval_loss": 1.1518688201904297,
+      "eval_runtime": 220.9554,
+      "eval_samples_per_second": 452.58,
+      "eval_steps_per_second": 14.143,
+      "step": 314000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.048666504007057e-05,
+      "loss": 1.3183,
+      "step": 314500
+    },
+    {
+      "epoch": 2.95,
+      "eval_loss": 1.159049391746521,
+      "eval_runtime": 220.9401,
+      "eval_samples_per_second": 452.611,
+      "eval_steps_per_second": 14.144,
+      "step": 314500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.0439743998798823e-05,
+      "loss": 1.3662,
+      "step": 315000
+    },
+    {
+      "epoch": 2.96,
+      "eval_loss": 1.15578031539917,
+      "eval_runtime": 220.905,
+      "eval_samples_per_second": 452.683,
+      "eval_steps_per_second": 14.146,
+      "step": 315000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.0392822957527073e-05,
+      "loss": 1.3285,
+      "step": 315500
+    },
+    {
+      "epoch": 2.96,
+      "eval_loss": 1.1557625532150269,
+      "eval_runtime": 220.9213,
+      "eval_samples_per_second": 452.65,
+      "eval_steps_per_second": 14.145,
+      "step": 315500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.0345901916255326e-05,
+      "loss": 1.3375,
+      "step": 316000
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 1.153180480003357,
+      "eval_runtime": 220.9165,
+      "eval_samples_per_second": 452.66,
+      "eval_steps_per_second": 14.146,
+      "step": 316000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.029898087498358e-05,
+      "loss": 1.3208,
+      "step": 316500
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 1.1530534029006958,
+      "eval_runtime": 220.9013,
+      "eval_samples_per_second": 452.691,
+      "eval_steps_per_second": 14.147,
+      "step": 316500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.025205983371183e-05,
+      "loss": 1.3245,
+      "step": 317000
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 1.152458667755127,
+      "eval_runtime": 220.9087,
+      "eval_samples_per_second": 452.676,
+      "eval_steps_per_second": 14.146,
+      "step": 317000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.0205138792440083e-05,
+      "loss": 1.3383,
+      "step": 317500
+    },
+    {
+      "epoch": 2.98,
+      "eval_loss": 1.154115915298462,
+      "eval_runtime": 220.909,
+      "eval_samples_per_second": 452.675,
+      "eval_steps_per_second": 14.146,
+      "step": 317500
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.0158217751168336e-05,
+      "loss": 1.3139,
+      "step": 318000
+    },
+    {
+      "epoch": 2.98,
+      "eval_loss": 1.1526515483856201,
+      "eval_runtime": 220.8871,
+      "eval_samples_per_second": 452.72,
+      "eval_steps_per_second": 14.147,
+      "step": 318000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.011129670989659e-05,
+      "loss": 1.318,
+      "step": 318500
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 1.1521937847137451,
+      "eval_runtime": 220.928,
+      "eval_samples_per_second": 452.636,
+      "eval_steps_per_second": 14.145,
+      "step": 318500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.006437566862484e-05,
+      "loss": 1.3276,
+      "step": 319000
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 1.1527777910232544,
+      "eval_runtime": 220.9248,
+      "eval_samples_per_second": 452.643,
+      "eval_steps_per_second": 14.145,
+      "step": 319000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.0017454627353092e-05,
+      "loss": 1.3318,
+      "step": 319500
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.1514686346054077,
+      "eval_runtime": 220.8852,
+      "eval_samples_per_second": 452.724,
+      "eval_steps_per_second": 14.148,
+      "step": 319500
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.9970533586081346e-05,
+      "loss": 1.3396,
+      "step": 320000
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.153663158416748,
+      "eval_runtime": 220.9135,
+      "eval_samples_per_second": 452.666,
+      "eval_steps_per_second": 14.146,
+      "step": 320000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.9923612544809596e-05,
+      "loss": 1.3293,
+      "step": 320500
+    },
+    {
+      "epoch": 3.01,
+      "eval_loss": 1.1468321084976196,
+      "eval_runtime": 220.9867,
+      "eval_samples_per_second": 452.516,
+      "eval_steps_per_second": 14.141,
+      "step": 320500
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.987669150353785e-05,
+      "loss": 1.3166,
+      "step": 321000
+    },
+    {
+      "epoch": 3.01,
+      "eval_loss": 1.1455674171447754,
+      "eval_runtime": 221.0441,
+      "eval_samples_per_second": 452.398,
+      "eval_steps_per_second": 14.137,
+      "step": 321000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.9829770462266102e-05,
+      "loss": 1.3082,
+      "step": 321500
+    },
+    {
+      "epoch": 3.02,
+      "eval_loss": 1.1440938711166382,
+      "eval_runtime": 220.9927,
+      "eval_samples_per_second": 452.504,
+      "eval_steps_per_second": 14.141,
+      "step": 321500
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.9782849420994352e-05,
+      "loss": 1.3352,
+      "step": 322000
+    },
+    {
+      "epoch": 3.02,
+      "eval_loss": 1.1490601301193237,
+      "eval_runtime": 220.9673,
+      "eval_samples_per_second": 452.556,
+      "eval_steps_per_second": 14.142,
+      "step": 322000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.9735928379722605e-05,
+      "loss": 1.3217,
+      "step": 322500
+    },
+    {
+      "epoch": 3.03,
+      "eval_loss": 1.1430158615112305,
+      "eval_runtime": 220.8559,
+      "eval_samples_per_second": 452.784,
+      "eval_steps_per_second": 14.149,
+      "step": 322500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.968900733845086e-05,
+      "loss": 1.2891,
+      "step": 323000
+    },
+    {
+      "epoch": 3.03,
+      "eval_loss": 1.151960849761963,
+      "eval_runtime": 220.8197,
+      "eval_samples_per_second": 452.858,
+      "eval_steps_per_second": 14.152,
+      "step": 323000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.964208629717911e-05,
+      "loss": 1.3187,
+      "step": 323500
+    },
+    {
+      "epoch": 3.04,
+      "eval_loss": 1.1531556844711304,
+      "eval_runtime": 220.8223,
+      "eval_samples_per_second": 452.853,
+      "eval_steps_per_second": 14.152,
+      "step": 323500
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.9595165255907362e-05,
+      "loss": 1.3057,
+      "step": 324000
+    },
+    {
+      "epoch": 3.04,
+      "eval_loss": 1.1435818672180176,
+      "eval_runtime": 220.8506,
+      "eval_samples_per_second": 452.795,
+      "eval_steps_per_second": 14.15,
+      "step": 324000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9548244214635612e-05,
+      "loss": 1.3006,
+      "step": 324500
+    },
+    {
+      "epoch": 3.05,
+      "eval_loss": 1.1423208713531494,
+      "eval_runtime": 220.8327,
+      "eval_samples_per_second": 452.832,
+      "eval_steps_per_second": 14.151,
+      "step": 324500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9501323173363865e-05,
+      "loss": 1.3062,
+      "step": 325000
+    },
+    {
+      "epoch": 3.05,
+      "eval_loss": 1.1395783424377441,
+      "eval_runtime": 220.8318,
+      "eval_samples_per_second": 452.833,
+      "eval_steps_per_second": 14.151,
+      "step": 325000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.945440213209212e-05,
+      "loss": 1.2993,
+      "step": 325500
+    },
+    {
+      "epoch": 3.05,
+      "eval_loss": 1.139084815979004,
+      "eval_runtime": 220.827,
+      "eval_samples_per_second": 452.843,
+      "eval_steps_per_second": 14.151,
+      "step": 325500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.9407481090820368e-05,
+      "loss": 1.3258,
+      "step": 326000
+    },
+    {
+      "epoch": 3.06,
+      "eval_loss": 1.1458795070648193,
+      "eval_runtime": 220.8245,
+      "eval_samples_per_second": 452.848,
+      "eval_steps_per_second": 14.152,
+      "step": 326000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.936056004954862e-05,
+      "loss": 1.3166,
+      "step": 326500
+    },
+    {
+      "epoch": 3.06,
+      "eval_loss": 1.1415066719055176,
+      "eval_runtime": 220.8333,
+      "eval_samples_per_second": 452.83,
+      "eval_steps_per_second": 14.151,
+      "step": 326500
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.9313639008276875e-05,
+      "loss": 1.3062,
+      "step": 327000
+    },
+    {
+      "epoch": 3.07,
+      "eval_loss": 1.141654372215271,
+      "eval_runtime": 220.8651,
+      "eval_samples_per_second": 452.765,
+      "eval_steps_per_second": 14.149,
+      "step": 327000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.9266717967005125e-05,
+      "loss": 1.3091,
+      "step": 327500
+    },
+    {
+      "epoch": 3.07,
+      "eval_loss": 1.1461706161499023,
+      "eval_runtime": 220.8785,
+      "eval_samples_per_second": 452.738,
+      "eval_steps_per_second": 14.148,
+      "step": 327500
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.9219796925733378e-05,
+      "loss": 1.2958,
+      "step": 328000
+    },
+    {
+      "epoch": 3.08,
+      "eval_loss": 1.1399269104003906,
+      "eval_runtime": 220.8706,
+      "eval_samples_per_second": 452.754,
+      "eval_steps_per_second": 14.149,
+      "step": 328000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.917287588446163e-05,
+      "loss": 1.3067,
+      "step": 328500
+    },
+    {
+      "epoch": 3.08,
+      "eval_loss": 1.144376516342163,
+      "eval_runtime": 220.8285,
+      "eval_samples_per_second": 452.84,
+      "eval_steps_per_second": 14.151,
+      "step": 328500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.912595484318988e-05,
+      "loss": 1.319,
+      "step": 329000
+    },
+    {
+      "epoch": 3.09,
+      "eval_loss": 1.1402838230133057,
+      "eval_runtime": 220.844,
+      "eval_samples_per_second": 452.808,
+      "eval_steps_per_second": 14.15,
+      "step": 329000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.9079033801918134e-05,
+      "loss": 1.3399,
+      "step": 329500
+    },
+    {
+      "epoch": 3.09,
+      "eval_loss": 1.139337182044983,
+      "eval_runtime": 220.9399,
+      "eval_samples_per_second": 452.612,
+      "eval_steps_per_second": 14.144,
+      "step": 329500
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.9032112760646384e-05,
+      "loss": 1.3077,
+      "step": 330000
+    },
+    {
+      "epoch": 3.1,
+      "eval_loss": 1.1460225582122803,
+      "eval_runtime": 220.9227,
+      "eval_samples_per_second": 452.647,
+      "eval_steps_per_second": 14.145,
+      "step": 330000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.8985191719374638e-05,
+      "loss": 1.2905,
+      "step": 330500
+    },
+    {
+      "epoch": 3.1,
+      "eval_loss": 1.1372867822647095,
+      "eval_runtime": 220.8826,
+      "eval_samples_per_second": 452.729,
+      "eval_steps_per_second": 14.148,
+      "step": 330500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.893827067810289e-05,
+      "loss": 1.3308,
+      "step": 331000
+    },
+    {
+      "epoch": 3.11,
+      "eval_loss": 1.1457699537277222,
+      "eval_runtime": 220.9045,
+      "eval_samples_per_second": 452.684,
+      "eval_steps_per_second": 14.146,
+      "step": 331000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.889134963683114e-05,
+      "loss": 1.311,
+      "step": 331500
+    },
+    {
+      "epoch": 3.11,
+      "eval_loss": 1.1365880966186523,
+      "eval_runtime": 220.7213,
+      "eval_samples_per_second": 453.06,
+      "eval_steps_per_second": 14.158,
+      "step": 331500
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.8844428595559394e-05,
+      "loss": 1.3123,
+      "step": 332000
+    },
+    {
+      "epoch": 3.12,
+      "eval_loss": 1.1372052431106567,
+      "eval_runtime": 220.7014,
+      "eval_samples_per_second": 453.101,
+      "eval_steps_per_second": 14.159,
+      "step": 332000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.8797507554287647e-05,
+      "loss": 1.29,
+      "step": 332500
+    },
+    {
+      "epoch": 3.12,
+      "eval_loss": 1.1368725299835205,
+      "eval_runtime": 220.7248,
+      "eval_samples_per_second": 453.053,
+      "eval_steps_per_second": 14.158,
+      "step": 332500
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.8750586513015897e-05,
+      "loss": 1.2973,
+      "step": 333000
+    },
+    {
+      "epoch": 3.12,
+      "eval_loss": 1.1378732919692993,
+      "eval_runtime": 220.7445,
+      "eval_samples_per_second": 453.012,
+      "eval_steps_per_second": 14.157,
+      "step": 333000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.870366547174415e-05,
+      "loss": 1.294,
+      "step": 333500
+    },
+    {
+      "epoch": 3.13,
+      "eval_loss": 1.134841799736023,
+      "eval_runtime": 220.7851,
+      "eval_samples_per_second": 452.929,
+      "eval_steps_per_second": 14.154,
+      "step": 333500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.8656744430472404e-05,
+      "loss": 1.3215,
+      "step": 334000
+    },
+    {
+      "epoch": 3.13,
+      "eval_loss": 1.135761022567749,
+      "eval_runtime": 220.7626,
+      "eval_samples_per_second": 452.975,
+      "eval_steps_per_second": 14.155,
+      "step": 334000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.8609823389200654e-05,
+      "loss": 1.3083,
+      "step": 334500
+    },
+    {
+      "epoch": 3.14,
+      "eval_loss": 1.135439157485962,
+      "eval_runtime": 220.7917,
+      "eval_samples_per_second": 452.916,
+      "eval_steps_per_second": 14.154,
+      "step": 334500
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.8562902347928907e-05,
+      "loss": 1.2956,
+      "step": 335000
+    },
+    {
+      "epoch": 3.14,
+      "eval_loss": 1.1306705474853516,
+      "eval_runtime": 220.7752,
+      "eval_samples_per_second": 452.949,
+      "eval_steps_per_second": 14.155,
+      "step": 335000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.8515981306657157e-05,
+      "loss": 1.2909,
+      "step": 335500
+    },
+    {
+      "epoch": 3.15,
+      "eval_loss": 1.1373645067214966,
+      "eval_runtime": 220.7747,
+      "eval_samples_per_second": 452.95,
+      "eval_steps_per_second": 14.155,
+      "step": 335500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.846906026538541e-05,
+      "loss": 1.3079,
+      "step": 336000
+    },
+    {
+      "epoch": 3.15,
+      "eval_loss": 1.1389541625976562,
+      "eval_runtime": 220.7343,
+      "eval_samples_per_second": 453.033,
+      "eval_steps_per_second": 14.157,
+      "step": 336000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.8422139224113664e-05,
+      "loss": 1.2948,
+      "step": 336500
+    },
+    {
+      "epoch": 3.16,
+      "eval_loss": 1.13786780834198,
+      "eval_runtime": 220.7342,
+      "eval_samples_per_second": 453.034,
+      "eval_steps_per_second": 14.157,
+      "step": 336500
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.8375218182841913e-05,
+      "loss": 1.287,
+      "step": 337000
+    },
+    {
+      "epoch": 3.16,
+      "eval_loss": 1.1371397972106934,
+      "eval_runtime": 220.7289,
+      "eval_samples_per_second": 453.044,
+      "eval_steps_per_second": 14.158,
+      "step": 337000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.8328297141570167e-05,
+      "loss": 1.3254,
+      "step": 337500
+    },
+    {
+      "epoch": 3.17,
+      "eval_loss": 1.1344914436340332,
+      "eval_runtime": 220.7461,
+      "eval_samples_per_second": 453.009,
+      "eval_steps_per_second": 14.157,
+      "step": 337500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.828137610029842e-05,
+      "loss": 1.2886,
+      "step": 338000
+    },
+    {
+      "epoch": 3.17,
+      "eval_loss": 1.1374002695083618,
+      "eval_runtime": 220.7432,
+      "eval_samples_per_second": 453.015,
+      "eval_steps_per_second": 14.157,
+      "step": 338000
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.823445505902667e-05,
+      "loss": 1.3021,
+      "step": 338500
+    },
+    {
+      "epoch": 3.18,
+      "eval_loss": 1.1403025388717651,
+      "eval_runtime": 220.7142,
+      "eval_samples_per_second": 453.075,
+      "eval_steps_per_second": 14.159,
+      "step": 338500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.8187534017754923e-05,
+      "loss": 1.2922,
+      "step": 339000
+    },
+    {
+      "epoch": 3.18,
+      "eval_loss": 1.1281777620315552,
+      "eval_runtime": 220.889,
+      "eval_samples_per_second": 452.716,
+      "eval_steps_per_second": 14.147,
+      "step": 339000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.8140612976483176e-05,
+      "loss": 1.305,
+      "step": 339500
+    },
+    {
+      "epoch": 3.19,
+      "eval_loss": 1.1293456554412842,
+      "eval_runtime": 220.8438,
+      "eval_samples_per_second": 452.809,
+      "eval_steps_per_second": 14.15,
+      "step": 339500
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.8093691935211426e-05,
+      "loss": 1.3056,
+      "step": 340000
+    },
+    {
+      "epoch": 3.19,
+      "eval_loss": 1.1341973543167114,
+      "eval_runtime": 220.8184,
+      "eval_samples_per_second": 452.861,
+      "eval_steps_per_second": 14.152,
+      "step": 340000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.804677089393968e-05,
+      "loss": 1.2914,
+      "step": 340500
+    },
+    {
+      "epoch": 3.2,
+      "eval_loss": 1.133408546447754,
+      "eval_runtime": 220.7019,
+      "eval_samples_per_second": 453.1,
+      "eval_steps_per_second": 14.159,
+      "step": 340500
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.799984985266793e-05,
+      "loss": 1.3,
+      "step": 341000
+    },
+    {
+      "epoch": 3.2,
+      "eval_loss": 1.1342601776123047,
+      "eval_runtime": 220.5902,
+      "eval_samples_per_second": 453.329,
+      "eval_steps_per_second": 14.167,
+      "step": 341000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.7952928811396183e-05,
+      "loss": 1.2899,
+      "step": 341500
+    },
+    {
+      "epoch": 3.2,
+      "eval_loss": 1.1248433589935303,
+      "eval_runtime": 220.5614,
+      "eval_samples_per_second": 453.388,
+      "eval_steps_per_second": 14.168,
+      "step": 341500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.7906007770124436e-05,
+      "loss": 1.315,
+      "step": 342000
+    },
+    {
+      "epoch": 3.21,
+      "eval_loss": 1.1279655694961548,
+      "eval_runtime": 220.57,
+      "eval_samples_per_second": 453.371,
+      "eval_steps_per_second": 14.168,
+      "step": 342000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.7859086728852686e-05,
+      "loss": 1.3153,
+      "step": 342500
+    },
+    {
+      "epoch": 3.21,
+      "eval_loss": 1.125070571899414,
+      "eval_runtime": 220.6119,
+      "eval_samples_per_second": 453.285,
+      "eval_steps_per_second": 14.165,
+      "step": 342500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.781216568758094e-05,
+      "loss": 1.3028,
+      "step": 343000
+    },
+    {
+      "epoch": 3.22,
+      "eval_loss": 1.1249561309814453,
+      "eval_runtime": 220.5892,
+      "eval_samples_per_second": 453.331,
+      "eval_steps_per_second": 14.167,
+      "step": 343000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.7765244646309193e-05,
+      "loss": 1.2917,
+      "step": 343500
+    },
+    {
+      "epoch": 3.22,
+      "eval_loss": 1.1268163919448853,
+      "eval_runtime": 220.6095,
+      "eval_samples_per_second": 453.29,
+      "eval_steps_per_second": 14.165,
+      "step": 343500
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.7718323605037442e-05,
+      "loss": 1.3129,
+      "step": 344000
+    },
+    {
+      "epoch": 3.23,
+      "eval_loss": 1.130949854850769,
+      "eval_runtime": 220.5805,
+      "eval_samples_per_second": 453.349,
+      "eval_steps_per_second": 14.167,
+      "step": 344000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.7671402563765696e-05,
+      "loss": 1.3006,
+      "step": 344500
+    },
+    {
+      "epoch": 3.23,
+      "eval_loss": 1.130674958229065,
+      "eval_runtime": 220.5791,
+      "eval_samples_per_second": 453.352,
+      "eval_steps_per_second": 14.167,
+      "step": 344500
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.762448152249395e-05,
+      "loss": 1.296,
+      "step": 345000
+    },
+    {
+      "epoch": 3.24,
+      "eval_loss": 1.1254706382751465,
+      "eval_runtime": 220.6122,
+      "eval_samples_per_second": 453.284,
+      "eval_steps_per_second": 14.165,
+      "step": 345000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.75775604812222e-05,
+      "loss": 1.2797,
+      "step": 345500
+    },
+    {
+      "epoch": 3.24,
+      "eval_loss": 1.1241164207458496,
+      "eval_runtime": 220.5527,
+      "eval_samples_per_second": 453.406,
+      "eval_steps_per_second": 14.169,
+      "step": 345500
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.7530639439950452e-05,
+      "loss": 1.2894,
+      "step": 346000
+    },
+    {
+      "epoch": 3.25,
+      "eval_loss": 1.1250615119934082,
+      "eval_runtime": 220.5648,
+      "eval_samples_per_second": 453.381,
+      "eval_steps_per_second": 14.168,
+      "step": 346000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.7483718398678706e-05,
+      "loss": 1.2935,
+      "step": 346500
+    },
+    {
+      "epoch": 3.25,
+      "eval_loss": 1.1238473653793335,
+      "eval_runtime": 220.5533,
+      "eval_samples_per_second": 453.405,
+      "eval_steps_per_second": 14.169,
+      "step": 346500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.7436797357406955e-05,
+      "loss": 1.2986,
+      "step": 347000
+    },
+    {
+      "epoch": 3.26,
+      "eval_loss": 1.1256763935089111,
+      "eval_runtime": 220.5765,
+      "eval_samples_per_second": 453.358,
+      "eval_steps_per_second": 14.167,
+      "step": 347000
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.738987631613521e-05,
+      "loss": 1.3051,
+      "step": 347500
+    },
+    {
+      "epoch": 3.26,
+      "eval_loss": 1.1246867179870605,
+      "eval_runtime": 220.5788,
+      "eval_samples_per_second": 453.353,
+      "eval_steps_per_second": 14.167,
+      "step": 347500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.734295527486346e-05,
+      "loss": 1.2825,
+      "step": 348000
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 1.1220121383666992,
+      "eval_runtime": 220.5841,
+      "eval_samples_per_second": 453.342,
+      "eval_steps_per_second": 14.167,
+      "step": 348000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.7296034233591712e-05,
+      "loss": 1.3064,
+      "step": 348500
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 1.125480055809021,
+      "eval_runtime": 220.6229,
+      "eval_samples_per_second": 453.262,
+      "eval_steps_per_second": 14.164,
+      "step": 348500
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.7249113192319965e-05,
+      "loss": 1.3057,
+      "step": 349000
+    },
+    {
+      "epoch": 3.28,
+      "eval_loss": 1.122577428817749,
+      "eval_runtime": 220.7794,
+      "eval_samples_per_second": 452.941,
+      "eval_steps_per_second": 14.154,
+      "step": 349000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.7202192151048215e-05,
+      "loss": 1.3045,
+      "step": 349500
+    },
+    {
+      "epoch": 3.28,
+      "eval_loss": 1.1199777126312256,
+      "eval_runtime": 220.7547,
+      "eval_samples_per_second": 452.991,
+      "eval_steps_per_second": 14.156,
+      "step": 349500
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.715527110977647e-05,
+      "loss": 1.2658,
+      "step": 350000
+    },
+    {
+      "epoch": 3.28,
+      "eval_loss": 1.1250447034835815,
+      "eval_runtime": 220.7117,
+      "eval_samples_per_second": 453.08,
+      "eval_steps_per_second": 14.159,
+      "step": 350000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.710835006850472e-05,
+      "loss": 1.2736,
+      "step": 350500
+    },
+    {
+      "epoch": 3.29,
+      "eval_loss": 1.1251693964004517,
+      "eval_runtime": 220.6204,
+      "eval_samples_per_second": 453.267,
+      "eval_steps_per_second": 14.165,
+      "step": 350500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.706142902723297e-05,
+      "loss": 1.3004,
+      "step": 351000
+    },
+    {
+      "epoch": 3.29,
+      "eval_loss": 1.121744155883789,
+      "eval_runtime": 220.47,
+      "eval_samples_per_second": 453.576,
+      "eval_steps_per_second": 14.174,
+      "step": 351000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.7014507985961225e-05,
+      "loss": 1.3014,
+      "step": 351500
+    },
+    {
+      "epoch": 3.3,
+      "eval_loss": 1.1183676719665527,
+      "eval_runtime": 220.4996,
+      "eval_samples_per_second": 453.516,
+      "eval_steps_per_second": 14.172,
+      "step": 351500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.6967586944689478e-05,
+      "loss": 1.2903,
+      "step": 352000
+    },
+    {
+      "epoch": 3.3,
+      "eval_loss": 1.128160834312439,
+      "eval_runtime": 220.4904,
+      "eval_samples_per_second": 453.534,
+      "eval_steps_per_second": 14.173,
+      "step": 352000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.6920665903417728e-05,
+      "loss": 1.2826,
+      "step": 352500
+    },
+    {
+      "epoch": 3.31,
+      "eval_loss": 1.1223293542861938,
+      "eval_runtime": 220.5271,
+      "eval_samples_per_second": 453.459,
+      "eval_steps_per_second": 14.171,
+      "step": 352500
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.687374486214598e-05,
+      "loss": 1.2968,
+      "step": 353000
+    },
+    {
+      "epoch": 3.31,
+      "eval_loss": 1.119235873222351,
+      "eval_runtime": 220.499,
+      "eval_samples_per_second": 453.517,
+      "eval_steps_per_second": 14.172,
+      "step": 353000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.682682382087423e-05,
+      "loss": 1.2715,
+      "step": 353500
+    },
+    {
+      "epoch": 3.32,
+      "eval_loss": 1.1275289058685303,
+      "eval_runtime": 220.5053,
+      "eval_samples_per_second": 453.504,
+      "eval_steps_per_second": 14.172,
+      "step": 353500
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.6779902779602484e-05,
+      "loss": 1.2998,
+      "step": 354000
+    },
+    {
+      "epoch": 3.32,
+      "eval_loss": 1.1185322999954224,
+      "eval_runtime": 220.5197,
+      "eval_samples_per_second": 453.474,
+      "eval_steps_per_second": 14.171,
+      "step": 354000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.6732981738330738e-05,
+      "loss": 1.2875,
+      "step": 354500
+    },
+    {
+      "epoch": 3.33,
+      "eval_loss": 1.116925597190857,
+      "eval_runtime": 220.5237,
+      "eval_samples_per_second": 453.466,
+      "eval_steps_per_second": 14.171,
+      "step": 354500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.6686060697058988e-05,
+      "loss": 1.277,
+      "step": 355000
+    },
+    {
+      "epoch": 3.33,
+      "eval_loss": 1.116443157196045,
+      "eval_runtime": 220.4994,
+      "eval_samples_per_second": 453.516,
+      "eval_steps_per_second": 14.172,
+      "step": 355000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.663913965578724e-05,
+      "loss": 1.2764,
+      "step": 355500
+    },
+    {
+      "epoch": 3.34,
+      "eval_loss": 1.1129635572433472,
+      "eval_runtime": 220.5171,
+      "eval_samples_per_second": 453.48,
+      "eval_steps_per_second": 14.171,
+      "step": 355500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.6592218614515494e-05,
+      "loss": 1.2677,
+      "step": 356000
+    },
+    {
+      "epoch": 3.34,
+      "eval_loss": 1.1151468753814697,
+      "eval_runtime": 220.4847,
+      "eval_samples_per_second": 453.546,
+      "eval_steps_per_second": 14.173,
+      "step": 356000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.6545297573243744e-05,
+      "loss": 1.284,
+      "step": 356500
+    },
+    {
+      "epoch": 3.35,
+      "eval_loss": 1.1204290390014648,
+      "eval_runtime": 220.4567,
+      "eval_samples_per_second": 453.604,
+      "eval_steps_per_second": 14.175,
+      "step": 356500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.6498376531971997e-05,
+      "loss": 1.2776,
+      "step": 357000
+    },
+    {
+      "epoch": 3.35,
+      "eval_loss": 1.1218979358673096,
+      "eval_runtime": 220.4675,
+      "eval_samples_per_second": 453.582,
+      "eval_steps_per_second": 14.174,
+      "step": 357000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.645145549070025e-05,
+      "loss": 1.2877,
+      "step": 357500
+    },
+    {
+      "epoch": 3.35,
+      "eval_loss": 1.1146986484527588,
+      "eval_runtime": 220.4757,
+      "eval_samples_per_second": 453.565,
+      "eval_steps_per_second": 14.174,
+      "step": 357500
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.64045344494285e-05,
+      "loss": 1.2732,
+      "step": 358000
+    },
+    {
+      "epoch": 3.36,
+      "eval_loss": 1.1204952001571655,
+      "eval_runtime": 220.4932,
+      "eval_samples_per_second": 453.529,
+      "eval_steps_per_second": 14.173,
+      "step": 358000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.6357613408156754e-05,
+      "loss": 1.2887,
+      "step": 358500
+    },
+    {
+      "epoch": 3.36,
+      "eval_loss": 1.113034963607788,
+      "eval_runtime": 220.5465,
+      "eval_samples_per_second": 453.419,
+      "eval_steps_per_second": 14.169,
+      "step": 358500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.6310692366885004e-05,
+      "loss": 1.273,
+      "step": 359000
+    },
+    {
+      "epoch": 3.37,
+      "eval_loss": 1.1138362884521484,
+      "eval_runtime": 220.5129,
+      "eval_samples_per_second": 453.488,
+      "eval_steps_per_second": 14.172,
+      "step": 359000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.6263771325613257e-05,
+      "loss": 1.2645,
+      "step": 359500
+    },
+    {
+      "epoch": 3.37,
+      "eval_loss": 1.1159985065460205,
+      "eval_runtime": 220.5661,
+      "eval_samples_per_second": 453.379,
+      "eval_steps_per_second": 14.168,
+      "step": 359500
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.621685028434151e-05,
+      "loss": 1.2909,
+      "step": 360000
+    },
+    {
+      "epoch": 3.38,
+      "eval_loss": 1.1123580932617188,
+      "eval_runtime": 220.6926,
+      "eval_samples_per_second": 453.119,
+      "eval_steps_per_second": 14.16,
+      "step": 360000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.616992924306976e-05,
+      "loss": 1.2871,
+      "step": 360500
+    },
+    {
+      "epoch": 3.38,
+      "eval_loss": 1.112741231918335,
+      "eval_runtime": 220.5711,
+      "eval_samples_per_second": 453.369,
+      "eval_steps_per_second": 14.168,
+      "step": 360500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.6123008201798014e-05,
+      "loss": 1.2848,
+      "step": 361000
+    },
+    {
+      "epoch": 3.39,
+      "eval_loss": 1.1092541217803955,
+      "eval_runtime": 220.558,
+      "eval_samples_per_second": 453.396,
+      "eval_steps_per_second": 14.169,
+      "step": 361000
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.6076087160526267e-05,
+      "loss": 1.2805,
+      "step": 361500
+    },
+    {
+      "epoch": 3.39,
+      "eval_loss": 1.1068326234817505,
+      "eval_runtime": 220.4445,
+      "eval_samples_per_second": 453.629,
+      "eval_steps_per_second": 14.176,
+      "step": 361500
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.602916611925452e-05,
+      "loss": 1.2747,
+      "step": 362000
+    },
+    {
+      "epoch": 3.4,
+      "eval_loss": 1.108471393585205,
+      "eval_runtime": 220.3945,
+      "eval_samples_per_second": 453.732,
+      "eval_steps_per_second": 14.179,
+      "step": 362000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.598224507798277e-05,
+      "loss": 1.286,
+      "step": 362500
+    },
+    {
+      "epoch": 3.4,
+      "eval_loss": 1.116951584815979,
+      "eval_runtime": 220.4488,
+      "eval_samples_per_second": 453.62,
+      "eval_steps_per_second": 14.176,
+      "step": 362500
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.5935324036711023e-05,
+      "loss": 1.2923,
+      "step": 363000
+    },
+    {
+      "epoch": 3.41,
+      "eval_loss": 1.106821060180664,
+      "eval_runtime": 220.4737,
+      "eval_samples_per_second": 453.569,
+      "eval_steps_per_second": 14.174,
+      "step": 363000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.5888402995439277e-05,
+      "loss": 1.2452,
+      "step": 363500
+    },
+    {
+      "epoch": 3.41,
+      "eval_loss": 1.1107667684555054,
+      "eval_runtime": 220.4687,
+      "eval_samples_per_second": 453.579,
+      "eval_steps_per_second": 14.174,
+      "step": 363500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.5841481954167526e-05,
+      "loss": 1.301,
+      "step": 364000
+    },
+    {
+      "epoch": 3.42,
+      "eval_loss": 1.103507161140442,
+      "eval_runtime": 220.5105,
+      "eval_samples_per_second": 453.493,
+      "eval_steps_per_second": 14.172,
+      "step": 364000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.579456091289578e-05,
+      "loss": 1.2933,
+      "step": 364500
+    },
+    {
+      "epoch": 3.42,
+      "eval_loss": 1.1143946647644043,
+      "eval_runtime": 220.552,
+      "eval_samples_per_second": 453.408,
+      "eval_steps_per_second": 14.169,
+      "step": 364500
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.5747639871624033e-05,
+      "loss": 1.2676,
+      "step": 365000
+    },
+    {
+      "epoch": 3.43,
+      "eval_loss": 1.1107447147369385,
+      "eval_runtime": 220.5769,
+      "eval_samples_per_second": 453.357,
+      "eval_steps_per_second": 14.167,
+      "step": 365000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.5700718830352283e-05,
+      "loss": 1.2726,
+      "step": 365500
+    },
+    {
+      "epoch": 3.43,
+      "eval_loss": 1.109850525856018,
+      "eval_runtime": 220.6362,
+      "eval_samples_per_second": 453.235,
+      "eval_steps_per_second": 14.164,
+      "step": 365500
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.5653797789080536e-05,
+      "loss": 1.2798,
+      "step": 366000
+    },
+    {
+      "epoch": 3.43,
+      "eval_loss": 1.1126877069473267,
+      "eval_runtime": 220.6282,
+      "eval_samples_per_second": 453.251,
+      "eval_steps_per_second": 14.164,
+      "step": 366000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.560687674780879e-05,
+      "loss": 1.297,
+      "step": 366500
+    },
+    {
+      "epoch": 3.44,
+      "eval_loss": 1.1066131591796875,
+      "eval_runtime": 220.6565,
+      "eval_samples_per_second": 453.193,
+      "eval_steps_per_second": 14.162,
+      "step": 366500
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.555995570653704e-05,
+      "loss": 1.2886,
+      "step": 367000
+    },
+    {
+      "epoch": 3.44,
+      "eval_loss": 1.1080513000488281,
+      "eval_runtime": 220.6411,
+      "eval_samples_per_second": 453.225,
+      "eval_steps_per_second": 14.163,
+      "step": 367000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.5513034665265293e-05,
+      "loss": 1.294,
+      "step": 367500
+    },
+    {
+      "epoch": 3.45,
+      "eval_loss": 1.109612226486206,
+      "eval_runtime": 220.6484,
+      "eval_samples_per_second": 453.21,
+      "eval_steps_per_second": 14.163,
+      "step": 367500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.5466113623993546e-05,
+      "loss": 1.2738,
+      "step": 368000
+    },
+    {
+      "epoch": 3.45,
+      "eval_loss": 1.099606990814209,
+      "eval_runtime": 220.7045,
+      "eval_samples_per_second": 453.095,
+      "eval_steps_per_second": 14.159,
+      "step": 368000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.54191925827218e-05,
+      "loss": 1.2877,
+      "step": 368500
+    },
+    {
+      "epoch": 3.46,
+      "eval_loss": 1.106638789176941,
+      "eval_runtime": 220.7202,
+      "eval_samples_per_second": 453.062,
+      "eval_steps_per_second": 14.158,
+      "step": 368500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.537227154145005e-05,
+      "loss": 1.276,
+      "step": 369000
+    },
+    {
+      "epoch": 3.46,
+      "eval_loss": 1.1104212999343872,
+      "eval_runtime": 220.7131,
+      "eval_samples_per_second": 453.077,
+      "eval_steps_per_second": 14.159,
+      "step": 369000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5325350500178302e-05,
+      "loss": 1.2741,
+      "step": 369500
+    },
+    {
+      "epoch": 3.47,
+      "eval_loss": 1.1078869104385376,
+      "eval_runtime": 220.7191,
+      "eval_samples_per_second": 453.065,
+      "eval_steps_per_second": 14.158,
+      "step": 369500
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5278429458906556e-05,
+      "loss": 1.2641,
+      "step": 370000
+    },
+    {
+      "epoch": 3.47,
+      "eval_loss": 1.1033143997192383,
+      "eval_runtime": 220.9613,
+      "eval_samples_per_second": 452.568,
+      "eval_steps_per_second": 14.143,
+      "step": 370000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5231508417634804e-05,
+      "loss": 1.2774,
+      "step": 370500
+    },
+    {
+      "epoch": 3.48,
+      "eval_loss": 1.1071778535842896,
+      "eval_runtime": 220.9875,
+      "eval_samples_per_second": 452.514,
+      "eval_steps_per_second": 14.141,
+      "step": 370500
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 1.5184587376363057e-05,
+      "loss": 1.2555,
+      "step": 371000
+    },
+    {
+      "epoch": 3.48,
+      "eval_loss": 1.1127955913543701,
+      "eval_runtime": 220.9818,
+      "eval_samples_per_second": 452.526,
+      "eval_steps_per_second": 14.141,
+      "step": 371000
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.5137666335091307e-05,
+      "loss": 1.2704,
+      "step": 371500
+    },
+    {
+      "epoch": 3.49,
+      "eval_loss": 1.1075456142425537,
+      "eval_runtime": 220.945,
+      "eval_samples_per_second": 452.601,
+      "eval_steps_per_second": 14.144,
+      "step": 371500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.509074529381956e-05,
+      "loss": 1.273,
+      "step": 372000
+    },
+    {
+      "epoch": 3.49,
+      "eval_loss": 1.0989867448806763,
+      "eval_runtime": 220.9281,
+      "eval_samples_per_second": 452.636,
+      "eval_steps_per_second": 14.145,
+      "step": 372000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.5043824252547814e-05,
+      "loss": 1.2788,
+      "step": 372500
+    },
+    {
+      "epoch": 3.5,
+      "eval_loss": 1.107438087463379,
+      "eval_runtime": 220.898,
+      "eval_samples_per_second": 452.698,
+      "eval_steps_per_second": 14.147,
+      "step": 372500
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.4996903211276064e-05,
+      "loss": 1.2729,
+      "step": 373000
+    },
+    {
+      "epoch": 3.5,
+      "eval_loss": 1.104393720626831,
+      "eval_runtime": 220.8922,
+      "eval_samples_per_second": 452.71,
+      "eval_steps_per_second": 14.147,
+      "step": 373000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.4949982170004317e-05,
+      "loss": 1.261,
+      "step": 373500
+    },
+    {
+      "epoch": 3.51,
+      "eval_loss": 1.1048295497894287,
+      "eval_runtime": 220.9264,
+      "eval_samples_per_second": 452.639,
+      "eval_steps_per_second": 14.145,
+      "step": 373500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.490306112873257e-05,
+      "loss": 1.2739,
+      "step": 374000
+    },
+    {
+      "epoch": 3.51,
+      "eval_loss": 1.1010278463363647,
+      "eval_runtime": 220.921,
+      "eval_samples_per_second": 452.651,
+      "eval_steps_per_second": 14.145,
+      "step": 374000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.4856140087460822e-05,
+      "loss": 1.288,
+      "step": 374500
+    },
+    {
+      "epoch": 3.51,
+      "eval_loss": 1.1028172969818115,
+      "eval_runtime": 220.9673,
+      "eval_samples_per_second": 452.556,
+      "eval_steps_per_second": 14.142,
+      "step": 374500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.4809219046189073e-05,
+      "loss": 1.2669,
+      "step": 375000
+    },
+    {
+      "epoch": 3.52,
+      "eval_loss": 1.1051478385925293,
+      "eval_runtime": 221.0246,
+      "eval_samples_per_second": 452.438,
+      "eval_steps_per_second": 14.139,
+      "step": 375000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.4762298004917327e-05,
+      "loss": 1.2715,
+      "step": 375500
+    },
+    {
+      "epoch": 3.52,
+      "eval_loss": 1.1000804901123047,
+      "eval_runtime": 221.0676,
+      "eval_samples_per_second": 452.35,
+      "eval_steps_per_second": 14.136,
+      "step": 375500
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.4715376963645578e-05,
+      "loss": 1.2728,
+      "step": 376000
+    },
+    {
+      "epoch": 3.53,
+      "eval_loss": 1.1016603708267212,
+      "eval_runtime": 221.0626,
+      "eval_samples_per_second": 452.361,
+      "eval_steps_per_second": 14.136,
+      "step": 376000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 1.466845592237383e-05,
+      "loss": 1.2625,
+      "step": 376500
+    },
+    {
+      "epoch": 3.53,
+      "eval_loss": 1.0998036861419678,
+      "eval_runtime": 221.0903,
+      "eval_samples_per_second": 452.304,
+      "eval_steps_per_second": 14.134,
+      "step": 376500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.4621534881102081e-05,
+      "loss": 1.2652,
+      "step": 377000
+    },
+    {
+      "epoch": 3.54,
+      "eval_loss": 1.1029491424560547,
+      "eval_runtime": 221.0807,
+      "eval_samples_per_second": 452.324,
+      "eval_steps_per_second": 14.135,
+      "step": 377000
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.4574613839830335e-05,
+      "loss": 1.2623,
+      "step": 377500
+    },
+    {
+      "epoch": 3.54,
+      "eval_loss": 1.103481411933899,
+      "eval_runtime": 221.1214,
+      "eval_samples_per_second": 452.24,
+      "eval_steps_per_second": 14.133,
+      "step": 377500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.4527692798558586e-05,
+      "loss": 1.2741,
+      "step": 378000
+    },
+    {
+      "epoch": 3.55,
+      "eval_loss": 1.099875569343567,
+      "eval_runtime": 221.1227,
+      "eval_samples_per_second": 452.238,
+      "eval_steps_per_second": 14.132,
+      "step": 378000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.4480771757286838e-05,
+      "loss": 1.2711,
+      "step": 378500
+    },
+    {
+      "epoch": 3.55,
+      "eval_loss": 1.1044048070907593,
+      "eval_runtime": 221.595,
+      "eval_samples_per_second": 451.274,
+      "eval_steps_per_second": 14.102,
+      "step": 378500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4433850716015091e-05,
+      "loss": 1.2755,
+      "step": 379000
+    },
+    {
+      "epoch": 3.56,
+      "eval_loss": 1.099456548690796,
+      "eval_runtime": 221.7605,
+      "eval_samples_per_second": 450.937,
+      "eval_steps_per_second": 14.092,
+      "step": 379000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4386929674743343e-05,
+      "loss": 1.277,
+      "step": 379500
+    },
+    {
+      "epoch": 3.56,
+      "eval_loss": 1.0986175537109375,
+      "eval_runtime": 221.6625,
+      "eval_samples_per_second": 451.136,
+      "eval_steps_per_second": 14.098,
+      "step": 379500
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4340008633471594e-05,
+      "loss": 1.2927,
+      "step": 380000
+    },
+    {
+      "epoch": 3.57,
+      "eval_loss": 1.0979562997817993,
+      "eval_runtime": 221.6264,
+      "eval_samples_per_second": 451.21,
+      "eval_steps_per_second": 14.1,
+      "step": 380000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.4293087592199848e-05,
+      "loss": 1.2719,
+      "step": 380500
+    },
+    {
+      "epoch": 3.57,
+      "eval_loss": 1.0883769989013672,
+      "eval_runtime": 221.2672,
+      "eval_samples_per_second": 451.942,
+      "eval_steps_per_second": 14.123,
+      "step": 380500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4246166550928101e-05,
+      "loss": 1.2681,
+      "step": 381000
+    },
+    {
+      "epoch": 3.58,
+      "eval_loss": 1.0984450578689575,
+      "eval_runtime": 221.3116,
+      "eval_samples_per_second": 451.852,
+      "eval_steps_per_second": 14.12,
+      "step": 381000
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.419924550965635e-05,
+      "loss": 1.2686,
+      "step": 381500
+    },
+    {
+      "epoch": 3.58,
+      "eval_loss": 1.0925695896148682,
+      "eval_runtime": 221.3252,
+      "eval_samples_per_second": 451.824,
+      "eval_steps_per_second": 14.119,
+      "step": 381500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4152324468384604e-05,
+      "loss": 1.2669,
+      "step": 382000
+    },
+    {
+      "epoch": 3.58,
+      "eval_loss": 1.1006863117218018,
+      "eval_runtime": 221.3717,
+      "eval_samples_per_second": 451.729,
+      "eval_steps_per_second": 14.117,
+      "step": 382000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4105403427112854e-05,
+      "loss": 1.2682,
+      "step": 382500
+    },
+    {
+      "epoch": 3.59,
+      "eval_loss": 1.0927406549453735,
+      "eval_runtime": 221.3896,
+      "eval_samples_per_second": 451.692,
+      "eval_steps_per_second": 14.115,
+      "step": 382500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.4058482385841107e-05,
+      "loss": 1.2713,
+      "step": 383000
+    },
+    {
+      "epoch": 3.59,
+      "eval_loss": 1.0991029739379883,
+      "eval_runtime": 221.3668,
+      "eval_samples_per_second": 451.739,
+      "eval_steps_per_second": 14.117,
+      "step": 383000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.401156134456936e-05,
+      "loss": 1.2405,
+      "step": 383500
+    },
+    {
+      "epoch": 3.6,
+      "eval_loss": 1.1023343801498413,
+      "eval_runtime": 221.3768,
+      "eval_samples_per_second": 451.718,
+      "eval_steps_per_second": 14.116,
+      "step": 383500
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.396464030329761e-05,
+      "loss": 1.2564,
+      "step": 384000
+    },
+    {
+      "epoch": 3.6,
+      "eval_loss": 1.0970436334609985,
+      "eval_runtime": 221.4221,
+      "eval_samples_per_second": 451.626,
+      "eval_steps_per_second": 14.113,
+      "step": 384000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.3917719262025864e-05,
+      "loss": 1.2774,
+      "step": 384500
+    },
+    {
+      "epoch": 3.61,
+      "eval_loss": 1.0976711511611938,
+      "eval_runtime": 221.4388,
+      "eval_samples_per_second": 451.592,
+      "eval_steps_per_second": 14.112,
+      "step": 384500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.3870798220754117e-05,
+      "loss": 1.2732,
+      "step": 385000
+    },
+    {
+      "epoch": 3.61,
+      "eval_loss": 1.091884732246399,
+      "eval_runtime": 221.434,
+      "eval_samples_per_second": 451.602,
+      "eval_steps_per_second": 14.113,
+      "step": 385000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.3823877179482367e-05,
+      "loss": 1.2868,
+      "step": 385500
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 1.0924078226089478,
+      "eval_runtime": 221.4826,
+      "eval_samples_per_second": 451.503,
+      "eval_steps_per_second": 14.109,
+      "step": 385500
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.377695613821062e-05,
+      "loss": 1.2648,
+      "step": 386000
+    },
+    {
+      "epoch": 3.62,
+      "eval_loss": 1.0871402025222778,
+      "eval_runtime": 221.6939,
+      "eval_samples_per_second": 451.072,
+      "eval_steps_per_second": 14.096,
+      "step": 386000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.3730035096938873e-05,
+      "loss": 1.238,
+      "step": 386500
+    },
+    {
+      "epoch": 3.63,
+      "eval_loss": 1.0820244550704956,
+      "eval_runtime": 221.7091,
+      "eval_samples_per_second": 451.042,
+      "eval_steps_per_second": 14.095,
+      "step": 386500
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 1.3683114055667123e-05,
+      "loss": 1.261,
+      "step": 387000
+    },
+    {
+      "epoch": 3.63,
+      "eval_loss": 1.0901474952697754,
+      "eval_runtime": 221.6939,
+      "eval_samples_per_second": 451.072,
+      "eval_steps_per_second": 14.096,
+      "step": 387000
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.3636193014395377e-05,
+      "loss": 1.2589,
+      "step": 387500
+    },
+    {
+      "epoch": 3.64,
+      "eval_loss": 1.0906635522842407,
+      "eval_runtime": 221.6797,
+      "eval_samples_per_second": 451.101,
+      "eval_steps_per_second": 14.097,
+      "step": 387500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.358927197312363e-05,
+      "loss": 1.271,
+      "step": 388000
+    },
+    {
+      "epoch": 3.64,
+      "eval_loss": 1.0888662338256836,
+      "eval_runtime": 221.5473,
+      "eval_samples_per_second": 451.371,
+      "eval_steps_per_second": 14.105,
+      "step": 388000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.354235093185188e-05,
+      "loss": 1.261,
+      "step": 388500
+    },
+    {
+      "epoch": 3.65,
+      "eval_loss": 1.0955654382705688,
+      "eval_runtime": 221.4943,
+      "eval_samples_per_second": 451.479,
+      "eval_steps_per_second": 14.109,
+      "step": 388500
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.3495429890580133e-05,
+      "loss": 1.2545,
+      "step": 389000
+    },
+    {
+      "epoch": 3.65,
+      "eval_loss": 1.096941590309143,
+      "eval_runtime": 221.5197,
+      "eval_samples_per_second": 451.427,
+      "eval_steps_per_second": 14.107,
+      "step": 389000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.3448508849308383e-05,
+      "loss": 1.2533,
+      "step": 389500
+    },
+    {
+      "epoch": 3.66,
+      "eval_loss": 1.0853080749511719,
+      "eval_runtime": 221.5678,
+      "eval_samples_per_second": 451.329,
+      "eval_steps_per_second": 14.104,
+      "step": 389500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.3401587808036636e-05,
+      "loss": 1.2683,
+      "step": 390000
+    },
+    {
+      "epoch": 3.66,
+      "eval_loss": 1.0878088474273682,
+      "eval_runtime": 221.5397,
+      "eval_samples_per_second": 451.386,
+      "eval_steps_per_second": 14.106,
+      "step": 390000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.335466676676489e-05,
+      "loss": 1.2485,
+      "step": 390500
+    },
+    {
+      "epoch": 3.66,
+      "eval_loss": 1.0916895866394043,
+      "eval_runtime": 221.4993,
+      "eval_samples_per_second": 451.469,
+      "eval_steps_per_second": 14.108,
+      "step": 390500
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.330774572549314e-05,
+      "loss": 1.2733,
+      "step": 391000
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 1.085897445678711,
+      "eval_runtime": 221.5237,
+      "eval_samples_per_second": 451.419,
+      "eval_steps_per_second": 14.107,
+      "step": 391000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.3260824684221393e-05,
+      "loss": 1.269,
+      "step": 391500
+    },
+    {
+      "epoch": 3.67,
+      "eval_loss": 1.0844569206237793,
+      "eval_runtime": 221.5194,
+      "eval_samples_per_second": 451.428,
+      "eval_steps_per_second": 14.107,
+      "step": 391500
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.3213903642949646e-05,
+      "loss": 1.2515,
+      "step": 392000
+    },
+    {
+      "epoch": 3.68,
+      "eval_loss": 1.0949499607086182,
+      "eval_runtime": 221.5302,
+      "eval_samples_per_second": 451.406,
+      "eval_steps_per_second": 14.106,
+      "step": 392000
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 1.3166982601677896e-05,
+      "loss": 1.2674,
+      "step": 392500
+    },
+    {
+      "epoch": 3.68,
+      "eval_loss": 1.0817222595214844,
+      "eval_runtime": 221.5421,
+      "eval_samples_per_second": 451.382,
+      "eval_steps_per_second": 14.106,
+      "step": 392500
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.312006156040615e-05,
+      "loss": 1.2747,
+      "step": 393000
+    },
+    {
+      "epoch": 3.69,
+      "eval_loss": 1.0874230861663818,
+      "eval_runtime": 221.4987,
+      "eval_samples_per_second": 451.47,
+      "eval_steps_per_second": 14.108,
+      "step": 393000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.3073140519134402e-05,
+      "loss": 1.2679,
+      "step": 393500
+    },
+    {
+      "epoch": 3.69,
+      "eval_loss": 1.0829836130142212,
+      "eval_runtime": 221.764,
+      "eval_samples_per_second": 450.93,
+      "eval_steps_per_second": 14.092,
+      "step": 393500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.3026219477862652e-05,
+      "loss": 1.2376,
+      "step": 394000
+    },
+    {
+      "epoch": 3.7,
+      "eval_loss": 1.0813868045806885,
+      "eval_runtime": 221.788,
+      "eval_samples_per_second": 450.881,
+      "eval_steps_per_second": 14.09,
+      "step": 394000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.2979298436590906e-05,
+      "loss": 1.2527,
+      "step": 394500
+    },
+    {
+      "epoch": 3.7,
+      "eval_loss": 1.081449270248413,
+      "eval_runtime": 221.7504,
+      "eval_samples_per_second": 450.957,
+      "eval_steps_per_second": 14.092,
+      "step": 394500
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.2932377395319156e-05,
+      "loss": 1.2559,
+      "step": 395000
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 1.0823462009429932,
+      "eval_runtime": 221.6871,
+      "eval_samples_per_second": 451.086,
+      "eval_steps_per_second": 14.096,
+      "step": 395000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.2885456354047409e-05,
+      "loss": 1.2265,
+      "step": 395500
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 1.085395097732544,
+      "eval_runtime": 221.4734,
+      "eval_samples_per_second": 451.522,
+      "eval_steps_per_second": 14.11,
+      "step": 395500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.2838535312775662e-05,
+      "loss": 1.2799,
+      "step": 396000
+    },
+    {
+      "epoch": 3.72,
+      "eval_loss": 1.0858019590377808,
+      "eval_runtime": 221.5008,
+      "eval_samples_per_second": 451.466,
+      "eval_steps_per_second": 14.108,
+      "step": 396000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.2791614271503912e-05,
+      "loss": 1.2449,
+      "step": 396500
+    },
+    {
+      "epoch": 3.72,
+      "eval_loss": 1.0877788066864014,
+      "eval_runtime": 221.5121,
+      "eval_samples_per_second": 451.443,
+      "eval_steps_per_second": 14.108,
+      "step": 396500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2744693230232165e-05,
+      "loss": 1.2451,
+      "step": 397000
+    },
+    {
+      "epoch": 3.73,
+      "eval_loss": 1.0856256484985352,
+      "eval_runtime": 221.5418,
+      "eval_samples_per_second": 451.382,
+      "eval_steps_per_second": 14.106,
+      "step": 397000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2697772188960419e-05,
+      "loss": 1.2292,
+      "step": 397500
+    },
+    {
+      "epoch": 3.73,
+      "eval_loss": 1.080007791519165,
+      "eval_runtime": 221.5442,
+      "eval_samples_per_second": 451.377,
+      "eval_steps_per_second": 14.106,
+      "step": 397500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.2650851147688668e-05,
+      "loss": 1.2452,
+      "step": 398000
+    },
+    {
+      "epoch": 3.73,
+      "eval_loss": 1.0801745653152466,
+      "eval_runtime": 221.5032,
+      "eval_samples_per_second": 451.461,
+      "eval_steps_per_second": 14.108,
+      "step": 398000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.2603930106416922e-05,
+      "loss": 1.2385,
+      "step": 398500
+    },
+    {
+      "epoch": 3.74,
+      "eval_loss": 1.079880714416504,
+      "eval_runtime": 221.5406,
+      "eval_samples_per_second": 451.385,
+      "eval_steps_per_second": 14.106,
+      "step": 398500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.2557009065145175e-05,
+      "loss": 1.2576,
+      "step": 399000
+    },
+    {
+      "epoch": 3.74,
+      "eval_loss": 1.0792741775512695,
+      "eval_runtime": 221.5278,
+      "eval_samples_per_second": 451.411,
+      "eval_steps_per_second": 14.107,
+      "step": 399000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2510088023873427e-05,
+      "loss": 1.2567,
+      "step": 399500
+    },
+    {
+      "epoch": 3.75,
+      "eval_loss": 1.0734611749649048,
+      "eval_runtime": 221.555,
+      "eval_samples_per_second": 451.355,
+      "eval_steps_per_second": 14.105,
+      "step": 399500
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2463166982601678e-05,
+      "loss": 1.2384,
+      "step": 400000
+    },
+    {
+      "epoch": 3.75,
+      "eval_loss": 1.0803865194320679,
+      "eval_runtime": 221.5706,
+      "eval_samples_per_second": 451.324,
+      "eval_steps_per_second": 14.104,
+      "step": 400000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.241624594132993e-05,
+      "loss": 1.2782,
+      "step": 400500
+    },
+    {
+      "epoch": 3.76,
+      "eval_loss": 1.085302472114563,
+      "eval_runtime": 221.6737,
+      "eval_samples_per_second": 451.113,
+      "eval_steps_per_second": 14.097,
+      "step": 400500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.2369324900058183e-05,
+      "loss": 1.2418,
+      "step": 401000
+    },
+    {
+      "epoch": 3.76,
+      "eval_loss": 1.0813859701156616,
+      "eval_runtime": 221.7597,
+      "eval_samples_per_second": 450.938,
+      "eval_steps_per_second": 14.092,
+      "step": 401000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.2322403858786435e-05,
+      "loss": 1.2467,
+      "step": 401500
+    },
+    {
+      "epoch": 3.77,
+      "eval_loss": 1.0765790939331055,
+      "eval_runtime": 221.6965,
+      "eval_samples_per_second": 451.067,
+      "eval_steps_per_second": 14.096,
+      "step": 401500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.2275482817514688e-05,
+      "loss": 1.2557,
+      "step": 402000
+    },
+    {
+      "epoch": 3.77,
+      "eval_loss": 1.074906587600708,
+      "eval_runtime": 221.6682,
+      "eval_samples_per_second": 451.125,
+      "eval_steps_per_second": 14.098,
+      "step": 402000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.222856177624294e-05,
+      "loss": 1.2417,
+      "step": 402500
+    },
+    {
+      "epoch": 3.78,
+      "eval_loss": 1.0793219804763794,
+      "eval_runtime": 221.5416,
+      "eval_samples_per_second": 451.383,
+      "eval_steps_per_second": 14.106,
+      "step": 402500
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.2181640734971191e-05,
+      "loss": 1.2544,
+      "step": 403000
+    },
+    {
+      "epoch": 3.78,
+      "eval_loss": 1.0767314434051514,
+      "eval_runtime": 221.4973,
+      "eval_samples_per_second": 451.473,
+      "eval_steps_per_second": 14.109,
+      "step": 403000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.2134719693699444e-05,
+      "loss": 1.2453,
+      "step": 403500
+    },
+    {
+      "epoch": 3.79,
+      "eval_loss": 1.0776728391647339,
+      "eval_runtime": 221.5112,
+      "eval_samples_per_second": 451.444,
+      "eval_steps_per_second": 14.108,
+      "step": 403500
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.2087798652427696e-05,
+      "loss": 1.2508,
+      "step": 404000
+    },
+    {
+      "epoch": 3.79,
+      "eval_loss": 1.0749331712722778,
+      "eval_runtime": 221.5304,
+      "eval_samples_per_second": 451.405,
+      "eval_steps_per_second": 14.106,
+      "step": 404000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.2040877611155948e-05,
+      "loss": 1.2388,
+      "step": 404500
+    },
+    {
+      "epoch": 3.8,
+      "eval_loss": 1.079124093055725,
+      "eval_runtime": 221.5211,
+      "eval_samples_per_second": 451.424,
+      "eval_steps_per_second": 14.107,
+      "step": 404500
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.19939565698842e-05,
+      "loss": 1.2614,
+      "step": 405000
+    },
+    {
+      "epoch": 3.8,
+      "eval_loss": 1.0767608880996704,
+      "eval_runtime": 221.5647,
+      "eval_samples_per_second": 451.335,
+      "eval_steps_per_second": 14.104,
+      "step": 405000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.1947035528612453e-05,
+      "loss": 1.2355,
+      "step": 405500
+    },
+    {
+      "epoch": 3.81,
+      "eval_loss": 1.0779807567596436,
+      "eval_runtime": 221.5326,
+      "eval_samples_per_second": 451.401,
+      "eval_steps_per_second": 14.106,
+      "step": 405500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.1900114487340704e-05,
+      "loss": 1.2481,
+      "step": 406000
+    },
+    {
+      "epoch": 3.81,
+      "eval_loss": 1.076469898223877,
+      "eval_runtime": 221.5584,
+      "eval_samples_per_second": 451.348,
+      "eval_steps_per_second": 14.105,
+      "step": 406000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.1853193446068956e-05,
+      "loss": 1.2382,
+      "step": 406500
+    },
+    {
+      "epoch": 3.81,
+      "eval_loss": 1.0811249017715454,
+      "eval_runtime": 221.5716,
+      "eval_samples_per_second": 451.321,
+      "eval_steps_per_second": 14.104,
+      "step": 406500
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.1806272404797209e-05,
+      "loss": 1.2331,
+      "step": 407000
+    },
+    {
+      "epoch": 3.82,
+      "eval_loss": 1.0766884088516235,
+      "eval_runtime": 221.5762,
+      "eval_samples_per_second": 451.312,
+      "eval_steps_per_second": 14.104,
+      "step": 407000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 1.175935136352546e-05,
+      "loss": 1.235,
+      "step": 407500
+    },
+    {
+      "epoch": 3.82,
+      "eval_loss": 1.0687439441680908,
+      "eval_runtime": 221.5531,
+      "eval_samples_per_second": 451.359,
+      "eval_steps_per_second": 14.105,
+      "step": 407500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.1712430322253712e-05,
+      "loss": 1.2506,
+      "step": 408000
+    },
+    {
+      "epoch": 3.83,
+      "eval_loss": 1.0767359733581543,
+      "eval_runtime": 221.7729,
+      "eval_samples_per_second": 450.912,
+      "eval_steps_per_second": 14.091,
+      "step": 408000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.1665509280981964e-05,
+      "loss": 1.2455,
+      "step": 408500
+    },
+    {
+      "epoch": 3.83,
+      "eval_loss": 1.0734050273895264,
+      "eval_runtime": 221.7206,
+      "eval_samples_per_second": 451.018,
+      "eval_steps_per_second": 14.094,
+      "step": 408500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.1618588239710217e-05,
+      "loss": 1.235,
+      "step": 409000
+    },
+    {
+      "epoch": 3.84,
+      "eval_loss": 1.076144814491272,
+      "eval_runtime": 221.6417,
+      "eval_samples_per_second": 451.179,
+      "eval_steps_per_second": 14.099,
+      "step": 409000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.1571667198438469e-05,
+      "loss": 1.227,
+      "step": 409500
+    },
+    {
+      "epoch": 3.84,
+      "eval_loss": 1.072109341621399,
+      "eval_runtime": 221.5766,
+      "eval_samples_per_second": 451.311,
+      "eval_steps_per_second": 14.103,
+      "step": 409500
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.152474615716672e-05,
+      "loss": 1.2566,
+      "step": 410000
+    },
+    {
+      "epoch": 3.85,
+      "eval_loss": 1.0735670328140259,
+      "eval_runtime": 221.4239,
+      "eval_samples_per_second": 451.623,
+      "eval_steps_per_second": 14.113,
+      "step": 410000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.1477825115894974e-05,
+      "loss": 1.2523,
+      "step": 410500
+    },
+    {
+      "epoch": 3.85,
+      "eval_loss": 1.0737024545669556,
+      "eval_runtime": 221.4176,
+      "eval_samples_per_second": 451.635,
+      "eval_steps_per_second": 14.114,
+      "step": 410500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.1430904074623225e-05,
+      "loss": 1.239,
+      "step": 411000
+    },
+    {
+      "epoch": 3.86,
+      "eval_loss": 1.075139045715332,
+      "eval_runtime": 221.4292,
+      "eval_samples_per_second": 451.612,
+      "eval_steps_per_second": 14.113,
+      "step": 411000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.1383983033351477e-05,
+      "loss": 1.2632,
+      "step": 411500
+    },
+    {
+      "epoch": 3.86,
+      "eval_loss": 1.0686564445495605,
+      "eval_runtime": 221.4503,
+      "eval_samples_per_second": 451.569,
+      "eval_steps_per_second": 14.112,
+      "step": 411500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.1337061992079728e-05,
+      "loss": 1.2299,
+      "step": 412000
+    },
+    {
+      "epoch": 3.87,
+      "eval_loss": 1.0703197717666626,
+      "eval_runtime": 221.4314,
+      "eval_samples_per_second": 451.607,
+      "eval_steps_per_second": 14.113,
+      "step": 412000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 1.1290140950807982e-05,
+      "loss": 1.2451,
+      "step": 412500
+    },
+    {
+      "epoch": 3.87,
+      "eval_loss": 1.0710246562957764,
+      "eval_runtime": 221.4462,
+      "eval_samples_per_second": 451.577,
+      "eval_steps_per_second": 14.112,
+      "step": 412500
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.1243219909536233e-05,
+      "loss": 1.2372,
+      "step": 413000
+    },
+    {
+      "epoch": 3.88,
+      "eval_loss": 1.0705777406692505,
+      "eval_runtime": 221.433,
+      "eval_samples_per_second": 451.604,
+      "eval_steps_per_second": 14.113,
+      "step": 413000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.1196298868264485e-05,
+      "loss": 1.2282,
+      "step": 413500
+    },
+    {
+      "epoch": 3.88,
+      "eval_loss": 1.0712275505065918,
+      "eval_runtime": 221.4443,
+      "eval_samples_per_second": 451.581,
+      "eval_steps_per_second": 14.112,
+      "step": 413500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1149377826992736e-05,
+      "loss": 1.2334,
+      "step": 414000
+    },
+    {
+      "epoch": 3.89,
+      "eval_loss": 1.072120189666748,
+      "eval_runtime": 221.422,
+      "eval_samples_per_second": 451.626,
+      "eval_steps_per_second": 14.113,
+      "step": 414000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.110245678572099e-05,
+      "loss": 1.23,
+      "step": 414500
+    },
+    {
+      "epoch": 3.89,
+      "eval_loss": 1.0694152116775513,
+      "eval_runtime": 221.4396,
+      "eval_samples_per_second": 451.59,
+      "eval_steps_per_second": 14.112,
+      "step": 414500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1055535744449241e-05,
+      "loss": 1.2466,
+      "step": 415000
+    },
+    {
+      "epoch": 3.89,
+      "eval_loss": 1.0689135789871216,
+      "eval_runtime": 221.4928,
+      "eval_samples_per_second": 451.482,
+      "eval_steps_per_second": 14.109,
+      "step": 415000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.1008614703177493e-05,
+      "loss": 1.2166,
+      "step": 415500
+    },
+    {
+      "epoch": 3.9,
+      "eval_loss": 1.0654420852661133,
+      "eval_runtime": 221.667,
+      "eval_samples_per_second": 451.127,
+      "eval_steps_per_second": 14.098,
+      "step": 415500
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.0961693661905746e-05,
+      "loss": 1.242,
+      "step": 416000
+    },
+    {
+      "epoch": 3.9,
+      "eval_loss": 1.060294270515442,
+      "eval_runtime": 221.5802,
+      "eval_samples_per_second": 451.304,
+      "eval_steps_per_second": 14.103,
+      "step": 416000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.0914772620633998e-05,
+      "loss": 1.2328,
+      "step": 416500
+    },
+    {
+      "epoch": 3.91,
+      "eval_loss": 1.0750603675842285,
+      "eval_runtime": 221.529,
+      "eval_samples_per_second": 451.408,
+      "eval_steps_per_second": 14.107,
+      "step": 416500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.086785157936225e-05,
+      "loss": 1.2305,
+      "step": 417000
+    },
+    {
+      "epoch": 3.91,
+      "eval_loss": 1.0709779262542725,
+      "eval_runtime": 221.4321,
+      "eval_samples_per_second": 451.606,
+      "eval_steps_per_second": 14.113,
+      "step": 417000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.0820930538090501e-05,
+      "loss": 1.2583,
+      "step": 417500
+    },
+    {
+      "epoch": 3.92,
+      "eval_loss": 1.069564700126648,
+      "eval_runtime": 221.3517,
+      "eval_samples_per_second": 451.77,
+      "eval_steps_per_second": 14.118,
+      "step": 417500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.0774009496818754e-05,
+      "loss": 1.2281,
+      "step": 418000
+    },
+    {
+      "epoch": 3.92,
+      "eval_loss": 1.068252444267273,
+      "eval_runtime": 221.3357,
+      "eval_samples_per_second": 451.802,
+      "eval_steps_per_second": 14.119,
+      "step": 418000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.0727088455547006e-05,
+      "loss": 1.2311,
+      "step": 418500
+    },
+    {
+      "epoch": 3.93,
+      "eval_loss": 1.0643572807312012,
+      "eval_runtime": 221.2957,
+      "eval_samples_per_second": 451.884,
+      "eval_steps_per_second": 14.121,
+      "step": 418500
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.0680167414275257e-05,
+      "loss": 1.2295,
+      "step": 419000
+    },
+    {
+      "epoch": 3.93,
+      "eval_loss": 1.0733916759490967,
+      "eval_runtime": 221.3151,
+      "eval_samples_per_second": 451.844,
+      "eval_steps_per_second": 14.12,
+      "step": 419000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.063324637300351e-05,
+      "loss": 1.2443,
+      "step": 419500
+    },
+    {
+      "epoch": 3.94,
+      "eval_loss": 1.068743109703064,
+      "eval_runtime": 221.3228,
+      "eval_samples_per_second": 451.829,
+      "eval_steps_per_second": 14.12,
+      "step": 419500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.0586325331731762e-05,
+      "loss": 1.2457,
+      "step": 420000
+    },
+    {
+      "epoch": 3.94,
+      "eval_loss": 1.0654057264328003,
+      "eval_runtime": 221.3124,
+      "eval_samples_per_second": 451.85,
+      "eval_steps_per_second": 14.12,
+      "step": 420000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0539404290460014e-05,
+      "loss": 1.2431,
+      "step": 420500
+    },
+    {
+      "epoch": 3.95,
+      "eval_loss": 1.0636610984802246,
+      "eval_runtime": 221.147,
+      "eval_samples_per_second": 452.188,
+      "eval_steps_per_second": 14.131,
+      "step": 420500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0492483249188265e-05,
+      "loss": 1.2248,
+      "step": 421000
+    },
+    {
+      "epoch": 3.95,
+      "eval_loss": 1.0696020126342773,
+      "eval_runtime": 221.1217,
+      "eval_samples_per_second": 452.24,
+      "eval_steps_per_second": 14.132,
+      "step": 421000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.0445562207916519e-05,
+      "loss": 1.2361,
+      "step": 421500
+    },
+    {
+      "epoch": 3.96,
+      "eval_loss": 1.0683060884475708,
+      "eval_runtime": 221.1561,
+      "eval_samples_per_second": 452.169,
+      "eval_steps_per_second": 14.13,
+      "step": 421500
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.039864116664477e-05,
+      "loss": 1.2502,
+      "step": 422000
+    },
+    {
+      "epoch": 3.96,
+      "eval_loss": 1.0591408014297485,
+      "eval_runtime": 221.1321,
+      "eval_samples_per_second": 452.218,
+      "eval_steps_per_second": 14.132,
+      "step": 422000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.0351720125373022e-05,
+      "loss": 1.2188,
+      "step": 422500
+    },
+    {
+      "epoch": 3.96,
+      "eval_loss": 1.0668104887008667,
+      "eval_runtime": 221.1228,
+      "eval_samples_per_second": 452.237,
+      "eval_steps_per_second": 14.132,
+      "step": 422500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.0304799084101273e-05,
+      "loss": 1.2439,
+      "step": 423000
+    },
+    {
+      "epoch": 3.97,
+      "eval_loss": 1.06576406955719,
+      "eval_runtime": 221.2975,
+      "eval_samples_per_second": 451.88,
+      "eval_steps_per_second": 14.121,
+      "step": 423000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.0257878042829527e-05,
+      "loss": 1.242,
+      "step": 423500
+    },
+    {
+      "epoch": 3.97,
+      "eval_loss": 1.0613994598388672,
+      "eval_runtime": 221.2789,
+      "eval_samples_per_second": 451.918,
+      "eval_steps_per_second": 14.122,
+      "step": 423500
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0210957001557778e-05,
+      "loss": 1.2245,
+      "step": 424000
+    },
+    {
+      "epoch": 3.98,
+      "eval_loss": 1.0643956661224365,
+      "eval_runtime": 221.2185,
+      "eval_samples_per_second": 452.042,
+      "eval_steps_per_second": 14.126,
+      "step": 424000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0164035960286032e-05,
+      "loss": 1.2288,
+      "step": 424500
+    },
+    {
+      "epoch": 3.98,
+      "eval_loss": 1.0604496002197266,
+      "eval_runtime": 221.1793,
+      "eval_samples_per_second": 452.122,
+      "eval_steps_per_second": 14.129,
+      "step": 424500
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.0117114919014283e-05,
+      "loss": 1.2308,
+      "step": 425000
+    },
+    {
+      "epoch": 3.99,
+      "eval_loss": 1.0600295066833496,
+      "eval_runtime": 221.044,
+      "eval_samples_per_second": 452.399,
+      "eval_steps_per_second": 14.137,
+      "step": 425000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.0070193877742535e-05,
+      "loss": 1.2294,
+      "step": 425500
+    },
+    {
+      "epoch": 3.99,
+      "eval_loss": 1.066137433052063,
+      "eval_runtime": 220.9979,
+      "eval_samples_per_second": 452.493,
+      "eval_steps_per_second": 14.14,
+      "step": 425500
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0023272836470788e-05,
+      "loss": 1.2377,
+      "step": 426000
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.0587615966796875,
+      "eval_runtime": 220.9922,
+      "eval_samples_per_second": 452.505,
+      "eval_steps_per_second": 14.141,
+      "step": 426000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.97635179519904e-06,
+      "loss": 1.219,
+      "step": 426500
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.0649616718292236,
+      "eval_runtime": 221.0888,
+      "eval_samples_per_second": 452.307,
+      "eval_steps_per_second": 14.135,
+      "step": 426500
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.929430753927293e-06,
+      "loss": 1.2135,
+      "step": 427000
+    },
+    {
+      "epoch": 4.01,
+      "eval_loss": 1.0579185485839844,
+      "eval_runtime": 221.0924,
+      "eval_samples_per_second": 452.3,
+      "eval_steps_per_second": 14.134,
+      "step": 427000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.882509712655545e-06,
+      "loss": 1.2194,
+      "step": 427500
+    },
+    {
+      "epoch": 4.01,
+      "eval_loss": 1.0547699928283691,
+      "eval_runtime": 221.1208,
+      "eval_samples_per_second": 452.241,
+      "eval_steps_per_second": 14.133,
+      "step": 427500
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.835588671383796e-06,
+      "loss": 1.2386,
+      "step": 428000
+    },
+    {
+      "epoch": 4.02,
+      "eval_loss": 1.0614423751831055,
+      "eval_runtime": 221.0918,
+      "eval_samples_per_second": 452.301,
+      "eval_steps_per_second": 14.134,
+      "step": 428000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.788667630112048e-06,
+      "loss": 1.2259,
+      "step": 428500
+    },
+    {
+      "epoch": 4.02,
+      "eval_loss": 1.0620990991592407,
+      "eval_runtime": 221.0818,
+      "eval_samples_per_second": 452.321,
+      "eval_steps_per_second": 14.135,
+      "step": 428500
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.741746588840301e-06,
+      "loss": 1.2263,
+      "step": 429000
+    },
+    {
+      "epoch": 4.03,
+      "eval_loss": 1.0643956661224365,
+      "eval_runtime": 221.0828,
+      "eval_samples_per_second": 452.319,
+      "eval_steps_per_second": 14.135,
+      "step": 429000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.694825547568553e-06,
+      "loss": 1.2305,
+      "step": 429500
+    },
+    {
+      "epoch": 4.03,
+      "eval_loss": 1.0544147491455078,
+      "eval_runtime": 221.1024,
+      "eval_samples_per_second": 452.279,
+      "eval_steps_per_second": 14.134,
+      "step": 429500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.647904506296804e-06,
+      "loss": 1.2101,
+      "step": 430000
+    },
+    {
+      "epoch": 4.04,
+      "eval_loss": 1.0544720888137817,
+      "eval_runtime": 221.1133,
+      "eval_samples_per_second": 452.257,
+      "eval_steps_per_second": 14.133,
+      "step": 430000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.600983465025057e-06,
+      "loss": 1.2334,
+      "step": 430500
+    },
+    {
+      "epoch": 4.04,
+      "eval_loss": 1.0552457571029663,
+      "eval_runtime": 220.9889,
+      "eval_samples_per_second": 452.511,
+      "eval_steps_per_second": 14.141,
+      "step": 430500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.554062423753309e-06,
+      "loss": 1.2204,
+      "step": 431000
+    },
+    {
+      "epoch": 4.04,
+      "eval_loss": 1.0604522228240967,
+      "eval_runtime": 221.1288,
+      "eval_samples_per_second": 452.225,
+      "eval_steps_per_second": 14.132,
+      "step": 431000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.50714138248156e-06,
+      "loss": 1.2375,
+      "step": 431500
+    },
+    {
+      "epoch": 4.05,
+      "eval_loss": 1.0569729804992676,
+      "eval_runtime": 221.1398,
+      "eval_samples_per_second": 452.203,
+      "eval_steps_per_second": 14.131,
+      "step": 431500
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.460220341209812e-06,
+      "loss": 1.2491,
+      "step": 432000
+    },
+    {
+      "epoch": 4.05,
+      "eval_loss": 1.052681803703308,
+      "eval_runtime": 221.0809,
+      "eval_samples_per_second": 452.323,
+      "eval_steps_per_second": 14.135,
+      "step": 432000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.413299299938066e-06,
+      "loss": 1.2265,
+      "step": 432500
+    },
+    {
+      "epoch": 4.06,
+      "eval_loss": 1.057190179824829,
+      "eval_runtime": 221.0473,
+      "eval_samples_per_second": 452.392,
+      "eval_steps_per_second": 14.137,
+      "step": 432500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.366378258666317e-06,
+      "loss": 1.2213,
+      "step": 433000
+    },
+    {
+      "epoch": 4.06,
+      "eval_loss": 1.0613055229187012,
+      "eval_runtime": 220.9425,
+      "eval_samples_per_second": 452.606,
+      "eval_steps_per_second": 14.144,
+      "step": 433000
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 9.319457217394569e-06,
+      "loss": 1.2242,
+      "step": 433500
+    },
+    {
+      "epoch": 4.07,
+      "eval_loss": 1.054632306098938,
+      "eval_runtime": 220.92,
+      "eval_samples_per_second": 452.652,
+      "eval_steps_per_second": 14.145,
+      "step": 433500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 9.272536176122822e-06,
+      "loss": 1.2335,
+      "step": 434000
+    },
+    {
+      "epoch": 4.07,
+      "eval_loss": 1.0627985000610352,
+      "eval_runtime": 220.9221,
+      "eval_samples_per_second": 452.648,
+      "eval_steps_per_second": 14.145,
+      "step": 434000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 9.225615134851074e-06,
+      "loss": 1.2306,
+      "step": 434500
+    },
+    {
+      "epoch": 4.08,
+      "eval_loss": 1.058109164237976,
+      "eval_runtime": 220.938,
+      "eval_samples_per_second": 452.616,
+      "eval_steps_per_second": 14.144,
+      "step": 434500
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 9.178694093579325e-06,
+      "loss": 1.2317,
+      "step": 435000
+    },
+    {
+      "epoch": 4.08,
+      "eval_loss": 1.060577392578125,
+      "eval_runtime": 220.957,
+      "eval_samples_per_second": 452.577,
+      "eval_steps_per_second": 14.143,
+      "step": 435000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 9.131773052307577e-06,
+      "loss": 1.2207,
+      "step": 435500
+    },
+    {
+      "epoch": 4.09,
+      "eval_loss": 1.0583726167678833,
+      "eval_runtime": 220.9147,
+      "eval_samples_per_second": 452.663,
+      "eval_steps_per_second": 14.146,
+      "step": 435500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 9.08485201103583e-06,
+      "loss": 1.2118,
+      "step": 436000
+    },
+    {
+      "epoch": 4.09,
+      "eval_loss": 1.0572487115859985,
+      "eval_runtime": 220.9401,
+      "eval_samples_per_second": 452.611,
+      "eval_steps_per_second": 14.144,
+      "step": 436000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 9.037930969764082e-06,
+      "loss": 1.2033,
+      "step": 436500
+    },
+    {
+      "epoch": 4.1,
+      "eval_loss": 1.0609925985336304,
+      "eval_runtime": 220.9828,
+      "eval_samples_per_second": 452.524,
+      "eval_steps_per_second": 14.141,
+      "step": 436500
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.991009928492333e-06,
+      "loss": 1.2224,
+      "step": 437000
+    },
+    {
+      "epoch": 4.1,
+      "eval_loss": 1.0533952713012695,
+      "eval_runtime": 220.933,
+      "eval_samples_per_second": 452.626,
+      "eval_steps_per_second": 14.145,
+      "step": 437000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.944088887220585e-06,
+      "loss": 1.223,
+      "step": 437500
+    },
+    {
+      "epoch": 4.11,
+      "eval_loss": 1.0515124797821045,
+      "eval_runtime": 220.9716,
+      "eval_samples_per_second": 452.547,
+      "eval_steps_per_second": 14.142,
+      "step": 437500
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.897167845948838e-06,
+      "loss": 1.2215,
+      "step": 438000
+    },
+    {
+      "epoch": 4.11,
+      "eval_loss": 1.0595310926437378,
+      "eval_runtime": 220.9271,
+      "eval_samples_per_second": 452.638,
+      "eval_steps_per_second": 14.145,
+      "step": 438000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.85024680467709e-06,
+      "loss": 1.2241,
+      "step": 438500
+    },
+    {
+      "epoch": 4.11,
+      "eval_loss": 1.055729627609253,
+      "eval_runtime": 220.9211,
+      "eval_samples_per_second": 452.65,
+      "eval_steps_per_second": 14.145,
+      "step": 438500
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.803325763405341e-06,
+      "loss": 1.1953,
+      "step": 439000
+    },
+    {
+      "epoch": 4.12,
+      "eval_loss": 1.0478371381759644,
+      "eval_runtime": 220.9454,
+      "eval_samples_per_second": 452.6,
+      "eval_steps_per_second": 14.144,
+      "step": 439000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.756404722133595e-06,
+      "loss": 1.2235,
+      "step": 439500
+    },
+    {
+      "epoch": 4.12,
+      "eval_loss": 1.05348801612854,
+      "eval_runtime": 221.0593,
+      "eval_samples_per_second": 452.367,
+      "eval_steps_per_second": 14.136,
+      "step": 439500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.709483680861846e-06,
+      "loss": 1.211,
+      "step": 440000
+    },
+    {
+      "epoch": 4.13,
+      "eval_loss": 1.0495777130126953,
+      "eval_runtime": 221.0334,
+      "eval_samples_per_second": 452.42,
+      "eval_steps_per_second": 14.138,
+      "step": 440000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.662562639590098e-06,
+      "loss": 1.2257,
+      "step": 440500
+    },
+    {
+      "epoch": 4.13,
+      "eval_loss": 1.056633472442627,
+      "eval_runtime": 220.9965,
+      "eval_samples_per_second": 452.496,
+      "eval_steps_per_second": 14.14,
+      "step": 440500
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.61564159831835e-06,
+      "loss": 1.2171,
+      "step": 441000
+    },
+    {
+      "epoch": 4.14,
+      "eval_loss": 1.0503853559494019,
+      "eval_runtime": 220.9748,
+      "eval_samples_per_second": 452.54,
+      "eval_steps_per_second": 14.142,
+      "step": 441000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.568720557046603e-06,
+      "loss": 1.2105,
+      "step": 441500
+    },
+    {
+      "epoch": 4.14,
+      "eval_loss": 1.0520578622817993,
+      "eval_runtime": 220.8941,
+      "eval_samples_per_second": 452.706,
+      "eval_steps_per_second": 14.147,
+      "step": 441500
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.521799515774854e-06,
+      "loss": 1.2045,
+      "step": 442000
+    },
+    {
+      "epoch": 4.15,
+      "eval_loss": 1.053121566772461,
+      "eval_runtime": 220.8544,
+      "eval_samples_per_second": 452.787,
+      "eval_steps_per_second": 14.15,
+      "step": 442000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.474878474503106e-06,
+      "loss": 1.205,
+      "step": 442500
+    },
+    {
+      "epoch": 4.15,
+      "eval_loss": 1.0605696439743042,
+      "eval_runtime": 220.8587,
+      "eval_samples_per_second": 452.778,
+      "eval_steps_per_second": 14.149,
+      "step": 442500
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.427957433231359e-06,
+      "loss": 1.2145,
+      "step": 443000
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 1.0511351823806763,
+      "eval_runtime": 220.8862,
+      "eval_samples_per_second": 452.722,
+      "eval_steps_per_second": 14.148,
+      "step": 443000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.38103639195961e-06,
+      "loss": 1.1973,
+      "step": 443500
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 1.0576379299163818,
+      "eval_runtime": 220.8821,
+      "eval_samples_per_second": 452.73,
+      "eval_steps_per_second": 14.148,
+      "step": 443500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.334115350687862e-06,
+      "loss": 1.2061,
+      "step": 444000
+    },
+    {
+      "epoch": 4.17,
+      "eval_loss": 1.0536248683929443,
+      "eval_runtime": 220.8693,
+      "eval_samples_per_second": 452.756,
+      "eval_steps_per_second": 14.149,
+      "step": 444000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.287194309416114e-06,
+      "loss": 1.2241,
+      "step": 444500
+    },
+    {
+      "epoch": 4.17,
+      "eval_loss": 1.0456198453903198,
+      "eval_runtime": 220.876,
+      "eval_samples_per_second": 452.743,
+      "eval_steps_per_second": 14.148,
+      "step": 444500
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.240273268144367e-06,
+      "loss": 1.1978,
+      "step": 445000
+    },
+    {
+      "epoch": 4.18,
+      "eval_loss": 1.0541229248046875,
+      "eval_runtime": 220.8941,
+      "eval_samples_per_second": 452.706,
+      "eval_steps_per_second": 14.147,
+      "step": 445000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.193352226872619e-06,
+      "loss": 1.2131,
+      "step": 445500
+    },
+    {
+      "epoch": 4.18,
+      "eval_loss": 1.050695538520813,
+      "eval_runtime": 220.8941,
+      "eval_samples_per_second": 452.706,
+      "eval_steps_per_second": 14.147,
+      "step": 445500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.14643118560087e-06,
+      "loss": 1.1941,
+      "step": 446000
+    },
+    {
+      "epoch": 4.19,
+      "eval_loss": 1.0486228466033936,
+      "eval_runtime": 220.8943,
+      "eval_samples_per_second": 452.705,
+      "eval_steps_per_second": 14.147,
+      "step": 446000
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.099510144329122e-06,
+      "loss": 1.2137,
+      "step": 446500
+    },
+    {
+      "epoch": 4.19,
+      "eval_loss": 1.050898790359497,
+      "eval_runtime": 220.9099,
+      "eval_samples_per_second": 452.673,
+      "eval_steps_per_second": 14.146,
+      "step": 446500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.052589103057375e-06,
+      "loss": 1.212,
+      "step": 447000
+    },
+    {
+      "epoch": 4.19,
+      "eval_loss": 1.052153468132019,
+      "eval_runtime": 220.9042,
+      "eval_samples_per_second": 452.685,
+      "eval_steps_per_second": 14.146,
+      "step": 447000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.005668061785627e-06,
+      "loss": 1.2141,
+      "step": 447500
+    },
+    {
+      "epoch": 4.2,
+      "eval_loss": 1.0504639148712158,
+      "eval_runtime": 220.8838,
+      "eval_samples_per_second": 452.727,
+      "eval_steps_per_second": 14.148,
+      "step": 447500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 7.958747020513878e-06,
+      "loss": 1.2035,
+      "step": 448000
+    },
+    {
+      "epoch": 4.2,
+      "eval_loss": 1.0515419244766235,
+      "eval_runtime": 220.9254,
+      "eval_samples_per_second": 452.642,
+      "eval_steps_per_second": 14.145,
+      "step": 448000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.911825979242132e-06,
+      "loss": 1.2162,
+      "step": 448500
+    },
+    {
+      "epoch": 4.21,
+      "eval_loss": 1.045407772064209,
+      "eval_runtime": 221.0538,
+      "eval_samples_per_second": 452.379,
+      "eval_steps_per_second": 14.137,
+      "step": 448500
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 7.864904937970383e-06,
+      "loss": 1.2062,
+      "step": 449000
+    },
+    {
+      "epoch": 4.21,
+      "eval_loss": 1.0419821739196777,
+      "eval_runtime": 221.0256,
+      "eval_samples_per_second": 452.436,
+      "eval_steps_per_second": 14.139,
+      "step": 449000
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 7.817983896698637e-06,
+      "loss": 1.1996,
+      "step": 449500
+    },
+    {
+      "epoch": 4.22,
+      "eval_loss": 1.0498727560043335,
+      "eval_runtime": 220.9307,
+      "eval_samples_per_second": 452.631,
+      "eval_steps_per_second": 14.145,
+      "step": 449500
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 7.771062855426888e-06,
+      "loss": 1.2194,
+      "step": 450000
+    },
+    {
+      "epoch": 4.22,
+      "eval_loss": 1.0517551898956299,
+      "eval_runtime": 220.9312,
+      "eval_samples_per_second": 452.63,
+      "eval_steps_per_second": 14.145,
+      "step": 450000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.72414181415514e-06,
+      "loss": 1.2062,
+      "step": 450500
+    },
+    {
+      "epoch": 4.23,
+      "eval_loss": 1.0490150451660156,
+      "eval_runtime": 220.7475,
+      "eval_samples_per_second": 453.006,
+      "eval_steps_per_second": 14.156,
+      "step": 450500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.677220772883393e-06,
+      "loss": 1.2046,
+      "step": 451000
+    },
+    {
+      "epoch": 4.23,
+      "eval_loss": 1.0536123514175415,
+      "eval_runtime": 220.7673,
+      "eval_samples_per_second": 452.966,
+      "eval_steps_per_second": 14.155,
+      "step": 451000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 7.630299731611645e-06,
+      "loss": 1.2012,
+      "step": 451500
+    },
+    {
+      "epoch": 4.24,
+      "eval_loss": 1.0506072044372559,
+      "eval_runtime": 220.7213,
+      "eval_samples_per_second": 453.06,
+      "eval_steps_per_second": 14.158,
+      "step": 451500
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 7.583378690339897e-06,
+      "loss": 1.2161,
+      "step": 452000
+    },
+    {
+      "epoch": 4.24,
+      "eval_loss": 1.0469845533370972,
+      "eval_runtime": 220.7698,
+      "eval_samples_per_second": 452.96,
+      "eval_steps_per_second": 14.155,
+      "step": 452000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 7.536457649068149e-06,
+      "loss": 1.2054,
+      "step": 452500
+    },
+    {
+      "epoch": 4.25,
+      "eval_loss": 1.0446325540542603,
+      "eval_runtime": 220.7759,
+      "eval_samples_per_second": 452.948,
+      "eval_steps_per_second": 14.155,
+      "step": 452500
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 7.4895366077964e-06,
+      "loss": 1.208,
+      "step": 453000
+    },
+    {
+      "epoch": 4.25,
+      "eval_loss": 1.0471805334091187,
+      "eval_runtime": 220.7642,
+      "eval_samples_per_second": 452.972,
+      "eval_steps_per_second": 14.155,
+      "step": 453000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.442615566524652e-06,
+      "loss": 1.1995,
+      "step": 453500
+    },
+    {
+      "epoch": 4.26,
+      "eval_loss": 1.0422724485397339,
+      "eval_runtime": 220.7863,
+      "eval_samples_per_second": 452.927,
+      "eval_steps_per_second": 14.154,
+      "step": 453500
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.395694525252905e-06,
+      "loss": 1.2153,
+      "step": 454000
+    },
+    {
+      "epoch": 4.26,
+      "eval_loss": 1.046839714050293,
+      "eval_runtime": 220.7559,
+      "eval_samples_per_second": 452.989,
+      "eval_steps_per_second": 14.156,
+      "step": 454000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.348773483981157e-06,
+      "loss": 1.1999,
+      "step": 454500
+    },
+    {
+      "epoch": 4.27,
+      "eval_loss": 1.0423718690872192,
+      "eval_runtime": 220.7317,
+      "eval_samples_per_second": 453.039,
+      "eval_steps_per_second": 14.157,
+      "step": 454500
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.301852442709409e-06,
+      "loss": 1.1919,
+      "step": 455000
+    },
+    {
+      "epoch": 4.27,
+      "eval_loss": 1.040187120437622,
+      "eval_runtime": 220.7387,
+      "eval_samples_per_second": 453.024,
+      "eval_steps_per_second": 14.157,
+      "step": 455000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.254931401437661e-06,
+      "loss": 1.1848,
+      "step": 455500
+    },
+    {
+      "epoch": 4.27,
+      "eval_loss": 1.0394939184188843,
+      "eval_runtime": 220.7554,
+      "eval_samples_per_second": 452.99,
+      "eval_steps_per_second": 14.156,
+      "step": 455500
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 7.208010360165913e-06,
+      "loss": 1.2012,
+      "step": 456000
+    },
+    {
+      "epoch": 4.28,
+      "eval_loss": 1.0482308864593506,
+      "eval_runtime": 220.7304,
+      "eval_samples_per_second": 453.041,
+      "eval_steps_per_second": 14.158,
+      "step": 456000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 7.161089318894166e-06,
+      "loss": 1.2028,
+      "step": 456500
+    },
+    {
+      "epoch": 4.28,
+      "eval_loss": 1.0419962406158447,
+      "eval_runtime": 220.7448,
+      "eval_samples_per_second": 453.012,
+      "eval_steps_per_second": 14.157,
+      "step": 456500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.114168277622417e-06,
+      "loss": 1.212,
+      "step": 457000
+    },
+    {
+      "epoch": 4.29,
+      "eval_loss": 1.0430151224136353,
+      "eval_runtime": 220.7318,
+      "eval_samples_per_second": 453.038,
+      "eval_steps_per_second": 14.157,
+      "step": 457000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.06724723635067e-06,
+      "loss": 1.1949,
+      "step": 457500
+    },
+    {
+      "epoch": 4.29,
+      "eval_loss": 1.047107219696045,
+      "eval_runtime": 220.7582,
+      "eval_samples_per_second": 452.984,
+      "eval_steps_per_second": 14.156,
+      "step": 457500
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 7.020326195078922e-06,
+      "loss": 1.2095,
+      "step": 458000
+    },
+    {
+      "epoch": 4.3,
+      "eval_loss": 1.0399595499038696,
+      "eval_runtime": 220.8799,
+      "eval_samples_per_second": 452.735,
+      "eval_steps_per_second": 14.148,
+      "step": 458000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 6.973405153807174e-06,
+      "loss": 1.2085,
+      "step": 458500
+    },
+    {
+      "epoch": 4.3,
+      "eval_loss": 1.0479100942611694,
+      "eval_runtime": 220.8879,
+      "eval_samples_per_second": 452.718,
+      "eval_steps_per_second": 14.147,
+      "step": 458500
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 6.926484112535425e-06,
+      "loss": 1.1901,
+      "step": 459000
+    },
+    {
+      "epoch": 4.31,
+      "eval_loss": 1.0417808294296265,
+      "eval_runtime": 220.8257,
+      "eval_samples_per_second": 452.846,
+      "eval_steps_per_second": 14.151,
+      "step": 459000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 6.8795630712636785e-06,
+      "loss": 1.2034,
+      "step": 459500
+    },
+    {
+      "epoch": 4.31,
+      "eval_loss": 1.0389989614486694,
+      "eval_runtime": 220.7906,
+      "eval_samples_per_second": 452.918,
+      "eval_steps_per_second": 14.154,
+      "step": 459500
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 6.83264202999193e-06,
+      "loss": 1.225,
+      "step": 460000
+    },
+    {
+      "epoch": 4.32,
+      "eval_loss": 1.0430644750595093,
+      "eval_runtime": 220.6771,
+      "eval_samples_per_second": 453.151,
+      "eval_steps_per_second": 14.161,
+      "step": 460000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 6.785720988720182e-06,
+      "loss": 1.1967,
+      "step": 460500
+    },
+    {
+      "epoch": 4.32,
+      "eval_loss": 1.0424212217330933,
+      "eval_runtime": 220.4736,
+      "eval_samples_per_second": 453.569,
+      "eval_steps_per_second": 14.174,
+      "step": 460500
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 6.738799947448435e-06,
+      "loss": 1.2059,
+      "step": 461000
+    },
+    {
+      "epoch": 4.33,
+      "eval_loss": 1.0408573150634766,
+      "eval_runtime": 220.4712,
+      "eval_samples_per_second": 453.574,
+      "eval_steps_per_second": 14.174,
+      "step": 461000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 6.691878906176687e-06,
+      "loss": 1.1939,
+      "step": 461500
+    },
+    {
+      "epoch": 4.33,
+      "eval_loss": 1.0424449443817139,
+      "eval_runtime": 220.5212,
+      "eval_samples_per_second": 453.471,
+      "eval_steps_per_second": 14.171,
+      "step": 461500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 6.644957864904938e-06,
+      "loss": 1.2027,
+      "step": 462000
+    },
+    {
+      "epoch": 4.34,
+      "eval_loss": 1.04022216796875,
+      "eval_runtime": 220.5052,
+      "eval_samples_per_second": 453.504,
+      "eval_steps_per_second": 14.172,
+      "step": 462000
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 6.59803682363319e-06,
+      "loss": 1.1745,
+      "step": 462500
+    },
+    {
+      "epoch": 4.34,
+      "eval_loss": 1.0380291938781738,
+      "eval_runtime": 220.5027,
+      "eval_samples_per_second": 453.509,
+      "eval_steps_per_second": 14.172,
+      "step": 462500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 6.551115782361443e-06,
+      "loss": 1.2155,
+      "step": 463000
+    },
+    {
+      "epoch": 4.34,
+      "eval_loss": 1.0391091108322144,
+      "eval_runtime": 220.4856,
+      "eval_samples_per_second": 453.544,
+      "eval_steps_per_second": 14.173,
+      "step": 463000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 6.504194741089695e-06,
+      "loss": 1.2033,
+      "step": 463500
+    },
+    {
+      "epoch": 4.35,
+      "eval_loss": 1.042190670967102,
+      "eval_runtime": 220.4876,
+      "eval_samples_per_second": 453.54,
+      "eval_steps_per_second": 14.173,
+      "step": 463500
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 6.457273699817946e-06,
+      "loss": 1.2064,
+      "step": 464000
+    },
+    {
+      "epoch": 4.35,
+      "eval_loss": 1.040285348892212,
+      "eval_runtime": 220.4809,
+      "eval_samples_per_second": 453.554,
+      "eval_steps_per_second": 14.174,
+      "step": 464000
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 6.410352658546198e-06,
+      "loss": 1.1943,
+      "step": 464500
+    },
+    {
+      "epoch": 4.36,
+      "eval_loss": 1.0418440103530884,
+      "eval_runtime": 220.482,
+      "eval_samples_per_second": 453.552,
+      "eval_steps_per_second": 14.173,
+      "step": 464500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 6.363431617274451e-06,
+      "loss": 1.1875,
+      "step": 465000
+    },
+    {
+      "epoch": 4.36,
+      "eval_loss": 1.0404266119003296,
+      "eval_runtime": 220.484,
+      "eval_samples_per_second": 453.548,
+      "eval_steps_per_second": 14.173,
+      "step": 465000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 6.316510576002703e-06,
+      "loss": 1.1978,
+      "step": 465500
+    },
+    {
+      "epoch": 4.37,
+      "eval_loss": 1.0388667583465576,
+      "eval_runtime": 220.5052,
+      "eval_samples_per_second": 453.504,
+      "eval_steps_per_second": 14.172,
+      "step": 465500
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 6.269589534730954e-06,
+      "loss": 1.1963,
+      "step": 466000
+    },
+    {
+      "epoch": 4.37,
+      "eval_loss": 1.039662480354309,
+      "eval_runtime": 220.4721,
+      "eval_samples_per_second": 453.572,
+      "eval_steps_per_second": 14.174,
+      "step": 466000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 6.222668493459207e-06,
+      "loss": 1.1991,
+      "step": 466500
+    },
+    {
+      "epoch": 4.38,
+      "eval_loss": 1.0323457717895508,
+      "eval_runtime": 220.4655,
+      "eval_samples_per_second": 453.586,
+      "eval_steps_per_second": 14.175,
+      "step": 466500
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 6.175747452187459e-06,
+      "loss": 1.2045,
+      "step": 467000
+    },
+    {
+      "epoch": 4.38,
+      "eval_loss": 1.0490866899490356,
+      "eval_runtime": 220.4714,
+      "eval_samples_per_second": 453.574,
+      "eval_steps_per_second": 14.174,
+      "step": 467000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 6.128826410915712e-06,
+      "loss": 1.1782,
+      "step": 467500
+    },
+    {
+      "epoch": 4.39,
+      "eval_loss": 1.0345863103866577,
+      "eval_runtime": 220.4724,
+      "eval_samples_per_second": 453.571,
+      "eval_steps_per_second": 14.174,
+      "step": 467500
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 6.081905369643963e-06,
+      "loss": 1.1832,
+      "step": 468000
+    },
+    {
+      "epoch": 4.39,
+      "eval_loss": 1.0386844873428345,
+      "eval_runtime": 220.5673,
+      "eval_samples_per_second": 453.376,
+      "eval_steps_per_second": 14.168,
+      "step": 468000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 6.034984328372216e-06,
+      "loss": 1.2035,
+      "step": 468500
+    },
+    {
+      "epoch": 4.4,
+      "eval_loss": 1.0390254259109497,
+      "eval_runtime": 220.603,
+      "eval_samples_per_second": 453.303,
+      "eval_steps_per_second": 14.166,
+      "step": 468500
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 5.988063287100468e-06,
+      "loss": 1.2036,
+      "step": 469000
+    },
+    {
+      "epoch": 4.4,
+      "eval_loss": 1.0352883338928223,
+      "eval_runtime": 220.4977,
+      "eval_samples_per_second": 453.52,
+      "eval_steps_per_second": 14.172,
+      "step": 469000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 5.94114224582872e-06,
+      "loss": 1.2,
+      "step": 469500
+    },
+    {
+      "epoch": 4.41,
+      "eval_loss": 1.03839111328125,
+      "eval_runtime": 220.4659,
+      "eval_samples_per_second": 453.585,
+      "eval_steps_per_second": 14.175,
+      "step": 469500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 5.894221204556972e-06,
+      "loss": 1.1869,
+      "step": 470000
+    },
+    {
+      "epoch": 4.41,
+      "eval_loss": 1.0423153638839722,
+      "eval_runtime": 220.3637,
+      "eval_samples_per_second": 453.795,
+      "eval_steps_per_second": 14.181,
+      "step": 470000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5.8473001632852245e-06,
+      "loss": 1.2012,
+      "step": 470500
+    },
+    {
+      "epoch": 4.42,
+      "eval_loss": 1.0392744541168213,
+      "eval_runtime": 220.3412,
+      "eval_samples_per_second": 453.842,
+      "eval_steps_per_second": 14.183,
+      "step": 470500
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5.800379122013476e-06,
+      "loss": 1.1912,
+      "step": 471000
+    },
+    {
+      "epoch": 4.42,
+      "eval_loss": 1.0390135049819946,
+      "eval_runtime": 220.3713,
+      "eval_samples_per_second": 453.78,
+      "eval_steps_per_second": 14.181,
+      "step": 471000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5.7534580807417286e-06,
+      "loss": 1.2062,
+      "step": 471500
+    },
+    {
+      "epoch": 4.42,
+      "eval_loss": 1.0349584817886353,
+      "eval_runtime": 220.3752,
+      "eval_samples_per_second": 453.771,
+      "eval_steps_per_second": 14.18,
+      "step": 471500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 5.70653703946998e-06,
+      "loss": 1.2079,
+      "step": 472000
+    },
+    {
+      "epoch": 4.43,
+      "eval_loss": 1.0324441194534302,
+      "eval_runtime": 220.391,
+      "eval_samples_per_second": 453.739,
+      "eval_steps_per_second": 14.179,
+      "step": 472000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 5.659615998198233e-06,
+      "loss": 1.1967,
+      "step": 472500
+    },
+    {
+      "epoch": 4.43,
+      "eval_loss": 1.0354297161102295,
+      "eval_runtime": 220.4046,
+      "eval_samples_per_second": 453.711,
+      "eval_steps_per_second": 14.178,
+      "step": 472500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 5.612694956926484e-06,
+      "loss": 1.194,
+      "step": 473000
+    },
+    {
+      "epoch": 4.44,
+      "eval_loss": 1.0334423780441284,
+      "eval_runtime": 220.39,
+      "eval_samples_per_second": 453.741,
+      "eval_steps_per_second": 14.179,
+      "step": 473000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 5.565773915654737e-06,
+      "loss": 1.1962,
+      "step": 473500
+    },
+    {
+      "epoch": 4.44,
+      "eval_loss": 1.0328459739685059,
+      "eval_runtime": 220.3907,
+      "eval_samples_per_second": 453.74,
+      "eval_steps_per_second": 14.179,
+      "step": 473500
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 5.518852874382988e-06,
+      "loss": 1.209,
+      "step": 474000
+    },
+    {
+      "epoch": 4.45,
+      "eval_loss": 1.0326672792434692,
+      "eval_runtime": 220.4416,
+      "eval_samples_per_second": 453.635,
+      "eval_steps_per_second": 14.176,
+      "step": 474000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 5.471931833111241e-06,
+      "loss": 1.1964,
+      "step": 474500
+    },
+    {
+      "epoch": 4.45,
+      "eval_loss": 1.0362826585769653,
+      "eval_runtime": 220.4411,
+      "eval_samples_per_second": 453.636,
+      "eval_steps_per_second": 14.176,
+      "step": 474500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 5.425010791839493e-06,
+      "loss": 1.1993,
+      "step": 475000
+    },
+    {
+      "epoch": 4.46,
+      "eval_loss": 1.0309995412826538,
+      "eval_runtime": 220.4439,
+      "eval_samples_per_second": 453.63,
+      "eval_steps_per_second": 14.176,
+      "step": 475000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 5.378089750567745e-06,
+      "loss": 1.2089,
+      "step": 475500
+    },
+    {
+      "epoch": 4.46,
+      "eval_loss": 1.0346201658248901,
+      "eval_runtime": 220.3965,
+      "eval_samples_per_second": 453.728,
+      "eval_steps_per_second": 14.179,
+      "step": 475500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5.331168709295997e-06,
+      "loss": 1.2001,
+      "step": 476000
+    },
+    {
+      "epoch": 4.47,
+      "eval_loss": 1.0281124114990234,
+      "eval_runtime": 220.429,
+      "eval_samples_per_second": 453.661,
+      "eval_steps_per_second": 14.177,
+      "step": 476000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5.284247668024249e-06,
+      "loss": 1.2331,
+      "step": 476500
+    },
+    {
+      "epoch": 4.47,
+      "eval_loss": 1.0332673788070679,
+      "eval_runtime": 220.4002,
+      "eval_samples_per_second": 453.72,
+      "eval_steps_per_second": 14.179,
+      "step": 476500
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.237326626752501e-06,
+      "loss": 1.2059,
+      "step": 477000
+    },
+    {
+      "epoch": 4.48,
+      "eval_loss": 1.033066987991333,
+      "eval_runtime": 220.4364,
+      "eval_samples_per_second": 453.646,
+      "eval_steps_per_second": 14.176,
+      "step": 477000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.190405585480753e-06,
+      "loss": 1.1919,
+      "step": 477500
+    },
+    {
+      "epoch": 4.48,
+      "eval_loss": 1.035326600074768,
+      "eval_runtime": 220.4207,
+      "eval_samples_per_second": 453.678,
+      "eval_steps_per_second": 14.177,
+      "step": 477500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5.143484544209005e-06,
+      "loss": 1.2028,
+      "step": 478000
+    },
+    {
+      "epoch": 4.49,
+      "eval_loss": 1.0330538749694824,
+      "eval_runtime": 220.3873,
+      "eval_samples_per_second": 453.747,
+      "eval_steps_per_second": 14.18,
+      "step": 478000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 5.096563502937257e-06,
+      "loss": 1.1815,
+      "step": 478500
+    },
+    {
+      "epoch": 4.49,
+      "eval_loss": 1.0284647941589355,
+      "eval_runtime": 220.3832,
+      "eval_samples_per_second": 453.755,
+      "eval_steps_per_second": 14.18,
+      "step": 478500
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5.049642461665509e-06,
+      "loss": 1.2108,
+      "step": 479000
+    },
+    {
+      "epoch": 4.5,
+      "eval_loss": 1.0274449586868286,
+      "eval_runtime": 220.5409,
+      "eval_samples_per_second": 453.431,
+      "eval_steps_per_second": 14.17,
+      "step": 479000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5.002721420393762e-06,
+      "loss": 1.1973,
+      "step": 479500
+    },
+    {
+      "epoch": 4.5,
+      "eval_loss": 1.030910849571228,
+      "eval_runtime": 220.517,
+      "eval_samples_per_second": 453.48,
+      "eval_steps_per_second": 14.171,
+      "step": 479500
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 4.955800379122014e-06,
+      "loss": 1.2209,
+      "step": 480000
+    },
+    {
+      "epoch": 4.5,
+      "eval_loss": 1.0277793407440186,
+      "eval_runtime": 220.48,
+      "eval_samples_per_second": 453.556,
+      "eval_steps_per_second": 14.174,
+      "step": 480000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 4.908879337850266e-06,
+      "loss": 1.1701,
+      "step": 480500
+    },
+    {
+      "epoch": 4.51,
+      "eval_loss": 1.0242763757705688,
+      "eval_runtime": 220.3969,
+      "eval_samples_per_second": 453.727,
+      "eval_steps_per_second": 14.179,
+      "step": 480500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 4.861958296578518e-06,
+      "loss": 1.1876,
+      "step": 481000
+    },
+    {
+      "epoch": 4.51,
+      "eval_loss": 1.0288127660751343,
+      "eval_runtime": 220.2841,
+      "eval_samples_per_second": 453.959,
+      "eval_steps_per_second": 14.186,
+      "step": 481000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 4.8150372553067706e-06,
+      "loss": 1.1949,
+      "step": 481500
+    },
+    {
+      "epoch": 4.52,
+      "eval_loss": 1.0324766635894775,
+      "eval_runtime": 220.2548,
+      "eval_samples_per_second": 454.02,
+      "eval_steps_per_second": 14.188,
+      "step": 481500
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 4.768116214035022e-06,
+      "loss": 1.1933,
+      "step": 482000
+    },
+    {
+      "epoch": 4.52,
+      "eval_loss": 1.0230944156646729,
+      "eval_runtime": 220.3139,
+      "eval_samples_per_second": 453.898,
+      "eval_steps_per_second": 14.184,
+      "step": 482000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 4.721195172763275e-06,
+      "loss": 1.1811,
+      "step": 482500
+    },
+    {
+      "epoch": 4.53,
+      "eval_loss": 1.029863953590393,
+      "eval_runtime": 220.2726,
+      "eval_samples_per_second": 453.983,
+      "eval_steps_per_second": 14.187,
+      "step": 482500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 4.674274131491526e-06,
+      "loss": 1.2033,
+      "step": 483000
+    },
+    {
+      "epoch": 4.53,
+      "eval_loss": 1.0319043397903442,
+      "eval_runtime": 220.2655,
+      "eval_samples_per_second": 453.998,
+      "eval_steps_per_second": 14.187,
+      "step": 483000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 4.627353090219779e-06,
+      "loss": 1.1919,
+      "step": 483500
+    },
+    {
+      "epoch": 4.54,
+      "eval_loss": 1.0290050506591797,
+      "eval_runtime": 220.268,
+      "eval_samples_per_second": 453.992,
+      "eval_steps_per_second": 14.187,
+      "step": 483500
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 4.58043204894803e-06,
+      "loss": 1.195,
+      "step": 484000
+    },
+    {
+      "epoch": 4.54,
+      "eval_loss": 1.0286130905151367,
+      "eval_runtime": 220.2866,
+      "eval_samples_per_second": 453.954,
+      "eval_steps_per_second": 14.186,
+      "step": 484000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.533511007676283e-06,
+      "loss": 1.1795,
+      "step": 484500
+    },
+    {
+      "epoch": 4.55,
+      "eval_loss": 1.0310746431350708,
+      "eval_runtime": 220.3028,
+      "eval_samples_per_second": 453.921,
+      "eval_steps_per_second": 14.185,
+      "step": 484500
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 4.486589966404535e-06,
+      "loss": 1.1981,
+      "step": 485000
+    },
+    {
+      "epoch": 4.55,
+      "eval_loss": 1.0278400182724,
+      "eval_runtime": 220.2765,
+      "eval_samples_per_second": 453.975,
+      "eval_steps_per_second": 14.187,
+      "step": 485000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.439668925132787e-06,
+      "loss": 1.1935,
+      "step": 485500
+    },
+    {
+      "epoch": 4.56,
+      "eval_loss": 1.0325404405593872,
+      "eval_runtime": 220.258,
+      "eval_samples_per_second": 454.013,
+      "eval_steps_per_second": 14.188,
+      "step": 485500
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 4.392747883861039e-06,
+      "loss": 1.1901,
+      "step": 486000
+    },
+    {
+      "epoch": 4.56,
+      "eval_loss": 1.029636263847351,
+      "eval_runtime": 220.2457,
+      "eval_samples_per_second": 454.038,
+      "eval_steps_per_second": 14.189,
+      "step": 486000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 4.345826842589291e-06,
+      "loss": 1.1998,
+      "step": 486500
+    },
+    {
+      "epoch": 4.57,
+      "eval_loss": 1.0245474576950073,
+      "eval_runtime": 220.2767,
+      "eval_samples_per_second": 453.975,
+      "eval_steps_per_second": 14.187,
+      "step": 486500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 4.298905801317543e-06,
+      "loss": 1.1818,
+      "step": 487000
+    },
+    {
+      "epoch": 4.57,
+      "eval_loss": 1.0290453433990479,
+      "eval_runtime": 220.2654,
+      "eval_samples_per_second": 453.998,
+      "eval_steps_per_second": 14.187,
+      "step": 487000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 4.251984760045795e-06,
+      "loss": 1.1713,
+      "step": 487500
+    },
+    {
+      "epoch": 4.57,
+      "eval_loss": 1.028717279434204,
+      "eval_runtime": 220.2563,
+      "eval_samples_per_second": 454.016,
+      "eval_steps_per_second": 14.188,
+      "step": 487500
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 4.205063718774047e-06,
+      "loss": 1.1841,
+      "step": 488000
+    },
+    {
+      "epoch": 4.58,
+      "eval_loss": 1.0243215560913086,
+      "eval_runtime": 220.2412,
+      "eval_samples_per_second": 454.048,
+      "eval_steps_per_second": 14.189,
+      "step": 488000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 4.158142677502299e-06,
+      "loss": 1.184,
+      "step": 488500
+    },
+    {
+      "epoch": 4.58,
+      "eval_loss": 1.0320249795913696,
+      "eval_runtime": 220.2122,
+      "eval_samples_per_second": 454.107,
+      "eval_steps_per_second": 14.191,
+      "step": 488500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 4.111221636230551e-06,
+      "loss": 1.1945,
+      "step": 489000
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 1.0234476327896118,
+      "eval_runtime": 220.2507,
+      "eval_samples_per_second": 454.028,
+      "eval_steps_per_second": 14.188,
+      "step": 489000
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 4.064300594958804e-06,
+      "loss": 1.1826,
+      "step": 489500
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 1.027974009513855,
+      "eval_runtime": 220.2473,
+      "eval_samples_per_second": 454.035,
+      "eval_steps_per_second": 14.189,
+      "step": 489500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 4.017379553687055e-06,
+      "loss": 1.188,
+      "step": 490000
+    },
+    {
+      "epoch": 4.6,
+      "eval_loss": 1.0277783870697021,
+      "eval_runtime": 220.2581,
+      "eval_samples_per_second": 454.013,
+      "eval_steps_per_second": 14.188,
+      "step": 490000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.970458512415308e-06,
+      "loss": 1.1939,
+      "step": 490500
+    },
+    {
+      "epoch": 4.6,
+      "eval_loss": 1.0183277130126953,
+      "eval_runtime": 220.2638,
+      "eval_samples_per_second": 454.001,
+      "eval_steps_per_second": 14.188,
+      "step": 490500
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.923537471143559e-06,
+      "loss": 1.1657,
+      "step": 491000
+    },
+    {
+      "epoch": 4.61,
+      "eval_loss": 1.029819369316101,
+      "eval_runtime": 220.2579,
+      "eval_samples_per_second": 454.013,
+      "eval_steps_per_second": 14.188,
+      "step": 491000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.876616429871812e-06,
+      "loss": 1.1917,
+      "step": 491500
+    },
+    {
+      "epoch": 4.61,
+      "eval_loss": 1.0290553569793701,
+      "eval_runtime": 220.4728,
+      "eval_samples_per_second": 453.571,
+      "eval_steps_per_second": 14.174,
+      "step": 491500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.829695388600064e-06,
+      "loss": 1.1945,
+      "step": 492000
+    },
+    {
+      "epoch": 4.62,
+      "eval_loss": 1.031397819519043,
+      "eval_runtime": 220.4389,
+      "eval_samples_per_second": 453.64,
+      "eval_steps_per_second": 14.176,
+      "step": 492000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.782774347328316e-06,
+      "loss": 1.2071,
+      "step": 492500
+    },
+    {
+      "epoch": 4.62,
+      "eval_loss": 1.027860164642334,
+      "eval_runtime": 220.3723,
+      "eval_samples_per_second": 453.778,
+      "eval_steps_per_second": 14.181,
+      "step": 492500
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 3.7358533060565677e-06,
+      "loss": 1.2013,
+      "step": 493000
+    },
+    {
+      "epoch": 4.63,
+      "eval_loss": 1.020934820175171,
+      "eval_runtime": 220.3406,
+      "eval_samples_per_second": 453.843,
+      "eval_steps_per_second": 14.183,
+      "step": 493000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 3.68893226478482e-06,
+      "loss": 1.1762,
+      "step": 493500
+    },
+    {
+      "epoch": 4.63,
+      "eval_loss": 1.0232270956039429,
+      "eval_runtime": 220.2361,
+      "eval_samples_per_second": 454.058,
+      "eval_steps_per_second": 14.189,
+      "step": 493500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.6420112235130726e-06,
+      "loss": 1.2014,
+      "step": 494000
+    },
+    {
+      "epoch": 4.64,
+      "eval_loss": 1.0220469236373901,
+      "eval_runtime": 220.2877,
+      "eval_samples_per_second": 453.952,
+      "eval_steps_per_second": 14.186,
+      "step": 494000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.595090182241324e-06,
+      "loss": 1.1817,
+      "step": 494500
+    },
+    {
+      "epoch": 4.64,
+      "eval_loss": 1.024511456489563,
+      "eval_runtime": 220.3417,
+      "eval_samples_per_second": 453.841,
+      "eval_steps_per_second": 14.183,
+      "step": 494500
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 3.5481691409695766e-06,
+      "loss": 1.1694,
+      "step": 495000
+    },
+    {
+      "epoch": 4.65,
+      "eval_loss": 1.0245423316955566,
+      "eval_runtime": 220.3569,
+      "eval_samples_per_second": 453.809,
+      "eval_steps_per_second": 14.182,
+      "step": 495000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 3.5012480996978287e-06,
+      "loss": 1.1864,
+      "step": 495500
+    },
+    {
+      "epoch": 4.65,
+      "eval_loss": 1.0235341787338257,
+      "eval_runtime": 220.3891,
+      "eval_samples_per_second": 453.743,
+      "eval_steps_per_second": 14.179,
+      "step": 495500
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 3.454327058426081e-06,
+      "loss": 1.2,
+      "step": 496000
+    },
+    {
+      "epoch": 4.65,
+      "eval_loss": 1.0256927013397217,
+      "eval_runtime": 220.4025,
+      "eval_samples_per_second": 453.715,
+      "eval_steps_per_second": 14.179,
+      "step": 496000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 3.4074060171543327e-06,
+      "loss": 1.1897,
+      "step": 496500
+    },
+    {
+      "epoch": 4.66,
+      "eval_loss": 1.0257229804992676,
+      "eval_runtime": 220.4604,
+      "eval_samples_per_second": 453.596,
+      "eval_steps_per_second": 14.175,
+      "step": 496500
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 3.360484975882585e-06,
+      "loss": 1.1929,
+      "step": 497000
+    },
+    {
+      "epoch": 4.66,
+      "eval_loss": 1.023755431175232,
+      "eval_runtime": 220.4521,
+      "eval_samples_per_second": 453.613,
+      "eval_steps_per_second": 14.175,
+      "step": 497000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 3.3135639346108367e-06,
+      "loss": 1.1853,
+      "step": 497500
+    },
+    {
+      "epoch": 4.67,
+      "eval_loss": 1.0197463035583496,
+      "eval_runtime": 220.4946,
+      "eval_samples_per_second": 453.526,
+      "eval_steps_per_second": 14.173,
+      "step": 497500
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 3.266642893339089e-06,
+      "loss": 1.1909,
+      "step": 498000
+    },
+    {
+      "epoch": 4.67,
+      "eval_loss": 1.0215179920196533,
+      "eval_runtime": 220.5236,
+      "eval_samples_per_second": 453.466,
+      "eval_steps_per_second": 14.171,
+      "step": 498000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 3.2197218520673416e-06,
+      "loss": 1.1784,
+      "step": 498500
+    },
+    {
+      "epoch": 4.68,
+      "eval_loss": 1.0229579210281372,
+      "eval_runtime": 220.8462,
+      "eval_samples_per_second": 452.804,
+      "eval_steps_per_second": 14.15,
+      "step": 498500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 3.172800810795593e-06,
+      "loss": 1.1955,
+      "step": 499000
+    },
+    {
+      "epoch": 4.68,
+      "eval_loss": 1.019065499305725,
+      "eval_runtime": 220.55,
+      "eval_samples_per_second": 453.412,
+      "eval_steps_per_second": 14.169,
+      "step": 499000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 3.1258797695238456e-06,
+      "loss": 1.177,
+      "step": 499500
+    },
+    {
+      "epoch": 4.69,
+      "eval_loss": 1.021791696548462,
+      "eval_runtime": 220.5844,
+      "eval_samples_per_second": 453.341,
+      "eval_steps_per_second": 14.167,
+      "step": 499500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 3.078958728252097e-06,
+      "loss": 1.1883,
+      "step": 500000
+    },
+    {
+      "epoch": 4.69,
+      "eval_loss": 1.0202616453170776,
+      "eval_runtime": 220.5578,
+      "eval_samples_per_second": 453.396,
+      "eval_steps_per_second": 14.169,
+      "step": 500000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 3.0320376869803496e-06,
+      "loss": 1.1911,
+      "step": 500500
+    },
+    {
+      "epoch": 4.7,
+      "eval_loss": 1.0237817764282227,
+      "eval_runtime": 220.6247,
+      "eval_samples_per_second": 453.258,
+      "eval_steps_per_second": 14.164,
+      "step": 500500
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 2.9851166457086017e-06,
+      "loss": 1.1923,
+      "step": 501000
+    },
+    {
+      "epoch": 4.7,
+      "eval_loss": 1.0245826244354248,
+      "eval_runtime": 220.6502,
+      "eval_samples_per_second": 453.206,
+      "eval_steps_per_second": 14.163,
+      "step": 501000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.9381956044368537e-06,
+      "loss": 1.1754,
+      "step": 501500
+    },
+    {
+      "epoch": 4.71,
+      "eval_loss": 1.0203845500946045,
+      "eval_runtime": 220.7296,
+      "eval_samples_per_second": 453.043,
+      "eval_steps_per_second": 14.158,
+      "step": 501500
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.891274563165106e-06,
+      "loss": 1.1848,
+      "step": 502000
+    },
+    {
+      "epoch": 4.71,
+      "eval_loss": 1.0270147323608398,
+      "eval_runtime": 221.2106,
+      "eval_samples_per_second": 452.058,
+      "eval_steps_per_second": 14.127,
+      "step": 502000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.844353521893358e-06,
+      "loss": 1.1739,
+      "step": 502500
+    },
+    {
+      "epoch": 4.72,
+      "eval_loss": 1.0190746784210205,
+      "eval_runtime": 221.5269,
+      "eval_samples_per_second": 451.412,
+      "eval_steps_per_second": 14.107,
+      "step": 502500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.79743248062161e-06,
+      "loss": 1.1925,
+      "step": 503000
+    },
+    {
+      "epoch": 4.72,
+      "eval_loss": 1.0178605318069458,
+      "eval_runtime": 221.4578,
+      "eval_samples_per_second": 451.553,
+      "eval_steps_per_second": 14.111,
+      "step": 503000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 2.750511439349862e-06,
+      "loss": 1.1799,
+      "step": 503500
+    },
+    {
+      "epoch": 4.72,
+      "eval_loss": 1.0173200368881226,
+      "eval_runtime": 221.4615,
+      "eval_samples_per_second": 451.546,
+      "eval_steps_per_second": 14.111,
+      "step": 503500
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.703590398078114e-06,
+      "loss": 1.1773,
+      "step": 504000
+    },
+    {
+      "epoch": 4.73,
+      "eval_loss": 1.0217922925949097,
+      "eval_runtime": 222.476,
+      "eval_samples_per_second": 449.487,
+      "eval_steps_per_second": 14.046,
+      "step": 504000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.656669356806366e-06,
+      "loss": 1.1675,
+      "step": 504500
+    },
+    {
+      "epoch": 4.73,
+      "eval_loss": 1.0150542259216309,
+      "eval_runtime": 242.0486,
+      "eval_samples_per_second": 413.14,
+      "eval_steps_per_second": 12.911,
+      "step": 504500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.6097483155346186e-06,
+      "loss": 1.1813,
+      "step": 505000
+    },
+    {
+      "epoch": 4.74,
+      "eval_loss": 1.0178701877593994,
+      "eval_runtime": 242.0285,
+      "eval_samples_per_second": 413.175,
+      "eval_steps_per_second": 12.912,
+      "step": 505000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.5628272742628706e-06,
+      "loss": 1.1726,
+      "step": 505500
+    },
+    {
+      "epoch": 4.74,
+      "eval_loss": 1.0183017253875732,
+      "eval_runtime": 242.2866,
+      "eval_samples_per_second": 412.734,
+      "eval_steps_per_second": 12.898,
+      "step": 505500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.5159062329911226e-06,
+      "loss": 1.1944,
+      "step": 506000
+    },
+    {
+      "epoch": 4.75,
+      "eval_loss": 1.011798620223999,
+      "eval_runtime": 242.0251,
+      "eval_samples_per_second": 413.18,
+      "eval_steps_per_second": 12.912,
+      "step": 506000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.4689851917193747e-06,
+      "loss": 1.1972,
+      "step": 506500
+    },
+    {
+      "epoch": 4.75,
+      "eval_loss": 1.021371841430664,
+      "eval_runtime": 242.3781,
+      "eval_samples_per_second": 412.579,
+      "eval_steps_per_second": 12.893,
+      "step": 506500
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.4220641504476267e-06,
+      "loss": 1.1835,
+      "step": 507000
+    },
+    {
+      "epoch": 4.76,
+      "eval_loss": 1.024190902709961,
+      "eval_runtime": 243.2084,
+      "eval_samples_per_second": 411.17,
+      "eval_steps_per_second": 12.849,
+      "step": 507000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 2.375143109175879e-06,
+      "loss": 1.1807,
+      "step": 507500
+    },
+    {
+      "epoch": 4.76,
+      "eval_loss": 1.0209712982177734,
+      "eval_runtime": 247.1697,
+      "eval_samples_per_second": 404.58,
+      "eval_steps_per_second": 12.643,
+      "step": 507500
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.328222067904131e-06,
+      "loss": 1.1801,
+      "step": 508000
+    },
+    {
+      "epoch": 4.77,
+      "eval_loss": 1.0197240114212036,
+      "eval_runtime": 245.9329,
+      "eval_samples_per_second": 406.615,
+      "eval_steps_per_second": 12.707,
+      "step": 508000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.281301026632383e-06,
+      "loss": 1.1707,
+      "step": 508500
+    },
+    {
+      "epoch": 4.77,
+      "eval_loss": 1.017072319984436,
+      "eval_runtime": 246.0676,
+      "eval_samples_per_second": 406.392,
+      "eval_steps_per_second": 12.7,
+      "step": 508500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.234379985360635e-06,
+      "loss": 1.1831,
+      "step": 509000
+    },
+    {
+      "epoch": 4.78,
+      "eval_loss": 1.0189136266708374,
+      "eval_runtime": 241.7862,
+      "eval_samples_per_second": 413.589,
+      "eval_steps_per_second": 12.925,
+      "step": 509000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 2.1874589440888876e-06,
+      "loss": 1.182,
+      "step": 509500
+    },
+    {
+      "epoch": 4.78,
+      "eval_loss": 1.0172502994537354,
+      "eval_runtime": 233.7568,
+      "eval_samples_per_second": 427.795,
+      "eval_steps_per_second": 13.369,
+      "step": 509500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.1405379028171396e-06,
+      "loss": 1.1946,
+      "step": 510000
+    },
+    {
+      "epoch": 4.79,
+      "eval_loss": 1.0199698209762573,
+      "eval_runtime": 233.9329,
+      "eval_samples_per_second": 427.473,
+      "eval_steps_per_second": 13.359,
+      "step": 510000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.0936168615453916e-06,
+      "loss": 1.1738,
+      "step": 510500
+    },
+    {
+      "epoch": 4.79,
+      "eval_loss": 1.016062617301941,
+      "eval_runtime": 235.9837,
+      "eval_samples_per_second": 423.758,
+      "eval_steps_per_second": 13.242,
+      "step": 510500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 2.0466958202736436e-06,
+      "loss": 1.1735,
+      "step": 511000
+    },
+    {
+      "epoch": 4.8,
+      "eval_loss": 1.0233088731765747,
+      "eval_runtime": 236.8841,
+      "eval_samples_per_second": 422.147,
+      "eval_steps_per_second": 13.192,
+      "step": 511000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.9997747790018957e-06,
+      "loss": 1.1771,
+      "step": 511500
+    },
+    {
+      "epoch": 4.8,
+      "eval_loss": 1.0144411325454712,
+      "eval_runtime": 236.4685,
+      "eval_samples_per_second": 422.889,
+      "eval_steps_per_second": 13.215,
+      "step": 511500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.9528537377301477e-06,
+      "loss": 1.165,
+      "step": 512000
+    },
+    {
+      "epoch": 4.8,
+      "eval_loss": 1.0135992765426636,
+      "eval_runtime": 233.2323,
+      "eval_samples_per_second": 428.757,
+      "eval_steps_per_second": 13.399,
+      "step": 512000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.9059326964583999e-06,
+      "loss": 1.1783,
+      "step": 512500
+    },
+    {
+      "epoch": 4.81,
+      "eval_loss": 1.0253207683563232,
+      "eval_runtime": 231.2367,
+      "eval_samples_per_second": 432.457,
+      "eval_steps_per_second": 13.514,
+      "step": 512500
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.859011655186652e-06,
+      "loss": 1.1783,
+      "step": 513000
+    },
+    {
+      "epoch": 4.81,
+      "eval_loss": 1.0126628875732422,
+      "eval_runtime": 233.9645,
+      "eval_samples_per_second": 427.415,
+      "eval_steps_per_second": 13.357,
+      "step": 513000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.812090613914904e-06,
+      "loss": 1.1824,
+      "step": 513500
+    },
+    {
+      "epoch": 4.82,
+      "eval_loss": 1.021931767463684,
+      "eval_runtime": 232.7488,
+      "eval_samples_per_second": 429.648,
+      "eval_steps_per_second": 13.426,
+      "step": 513500
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.765169572643156e-06,
+      "loss": 1.1553,
+      "step": 514000
+    },
+    {
+      "epoch": 4.82,
+      "eval_loss": 1.015001654624939,
+      "eval_runtime": 232.0708,
+      "eval_samples_per_second": 430.903,
+      "eval_steps_per_second": 13.466,
+      "step": 514000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.7182485313714084e-06,
+      "loss": 1.1799,
+      "step": 514500
+    },
+    {
+      "epoch": 4.83,
+      "eval_loss": 1.0170458555221558,
+      "eval_runtime": 231.7543,
+      "eval_samples_per_second": 431.491,
+      "eval_steps_per_second": 13.484,
+      "step": 514500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.6713274900996604e-06,
+      "loss": 1.1958,
+      "step": 515000
+    },
+    {
+      "epoch": 4.83,
+      "eval_loss": 1.0164183378219604,
+      "eval_runtime": 230.7078,
+      "eval_samples_per_second": 433.449,
+      "eval_steps_per_second": 13.545,
+      "step": 515000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.6244064488279126e-06,
+      "loss": 1.1946,
+      "step": 515500
+    },
+    {
+      "epoch": 4.84,
+      "eval_loss": 1.016738772392273,
+      "eval_runtime": 231.8187,
+      "eval_samples_per_second": 431.372,
+      "eval_steps_per_second": 13.48,
+      "step": 515500
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.5774854075561646e-06,
+      "loss": 1.1803,
+      "step": 516000
+    },
+    {
+      "epoch": 4.84,
+      "eval_loss": 1.0190505981445312,
+      "eval_runtime": 229.9801,
+      "eval_samples_per_second": 434.82,
+      "eval_steps_per_second": 13.588,
+      "step": 516000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.5305643662844166e-06,
+      "loss": 1.1723,
+      "step": 516500
+    },
+    {
+      "epoch": 4.85,
+      "eval_loss": 1.017173409461975,
+      "eval_runtime": 230.1856,
+      "eval_samples_per_second": 434.432,
+      "eval_steps_per_second": 13.576,
+      "step": 516500
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.4836433250126687e-06,
+      "loss": 1.1627,
+      "step": 517000
+    },
+    {
+      "epoch": 4.85,
+      "eval_loss": 1.013454556465149,
+      "eval_runtime": 229.3387,
+      "eval_samples_per_second": 436.036,
+      "eval_steps_per_second": 13.626,
+      "step": 517000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.4367222837409209e-06,
+      "loss": 1.1677,
+      "step": 517500
+    },
+    {
+      "epoch": 4.86,
+      "eval_loss": 1.0167146921157837,
+      "eval_runtime": 229.9528,
+      "eval_samples_per_second": 434.872,
+      "eval_steps_per_second": 13.59,
+      "step": 517500
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.389801242469173e-06,
+      "loss": 1.1923,
+      "step": 518000
+    },
+    {
+      "epoch": 4.86,
+      "eval_loss": 1.0137449502944946,
+      "eval_runtime": 233.5878,
+      "eval_samples_per_second": 428.105,
+      "eval_steps_per_second": 13.378,
+      "step": 518000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.3428802011974251e-06,
+      "loss": 1.1911,
+      "step": 518500
+    },
+    {
+      "epoch": 4.87,
+      "eval_loss": 1.0253547430038452,
+      "eval_runtime": 227.612,
+      "eval_samples_per_second": 439.344,
+      "eval_steps_per_second": 13.73,
+      "step": 518500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.2959591599256771e-06,
+      "loss": 1.1708,
+      "step": 519000
+    },
+    {
+      "epoch": 4.87,
+      "eval_loss": 1.018088936805725,
+      "eval_runtime": 229.1156,
+      "eval_samples_per_second": 436.461,
+      "eval_steps_per_second": 13.639,
+      "step": 519000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.2490381186539292e-06,
+      "loss": 1.1737,
+      "step": 519500
+    },
+    {
+      "epoch": 4.88,
+      "eval_loss": 1.0122038125991821,
+      "eval_runtime": 234.5523,
+      "eval_samples_per_second": 426.344,
+      "eval_steps_per_second": 13.323,
+      "step": 519500
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.2021170773821814e-06,
+      "loss": 1.1872,
+      "step": 520000
+    },
+    {
+      "epoch": 4.88,
+      "eval_loss": 1.0127085447311401,
+      "eval_runtime": 229.3109,
+      "eval_samples_per_second": 436.089,
+      "eval_steps_per_second": 13.628,
+      "step": 520000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.1551960361104334e-06,
+      "loss": 1.1734,
+      "step": 520500
+    },
+    {
+      "epoch": 4.88,
+      "eval_loss": 1.0173479318618774,
+      "eval_runtime": 228.8989,
+      "eval_samples_per_second": 436.874,
+      "eval_steps_per_second": 13.652,
+      "step": 520500
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.1082749948386854e-06,
+      "loss": 1.179,
+      "step": 521000
+    },
+    {
+      "epoch": 4.89,
+      "eval_loss": 1.0175164937973022,
+      "eval_runtime": 223.9589,
+      "eval_samples_per_second": 446.51,
+      "eval_steps_per_second": 13.953,
+      "step": 521000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.0613539535669376e-06,
+      "loss": 1.1778,
+      "step": 521500
+    },
+    {
+      "epoch": 4.89,
+      "eval_loss": 1.0166733264923096,
+      "eval_runtime": 229.4526,
+      "eval_samples_per_second": 435.82,
+      "eval_steps_per_second": 13.619,
+      "step": 521500
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.0144329122951896e-06,
+      "loss": 1.1732,
+      "step": 522000
+    },
+    {
+      "epoch": 4.9,
+      "eval_loss": 1.0115753412246704,
+      "eval_runtime": 229.4827,
+      "eval_samples_per_second": 435.763,
+      "eval_steps_per_second": 13.618,
+      "step": 522000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 9.675118710234419e-07,
+      "loss": 1.1635,
+      "step": 522500
+    },
+    {
+      "epoch": 4.9,
+      "eval_loss": 1.0163120031356812,
+      "eval_runtime": 227.8263,
+      "eval_samples_per_second": 438.931,
+      "eval_steps_per_second": 13.717,
+      "step": 522500
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 9.205908297516939e-07,
+      "loss": 1.1677,
+      "step": 523000
+    },
+    {
+      "epoch": 4.91,
+      "eval_loss": 1.0148788690567017,
+      "eval_runtime": 222.1269,
+      "eval_samples_per_second": 450.193,
+      "eval_steps_per_second": 14.069,
+      "step": 523000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 8.73669788479946e-07,
+      "loss": 1.1967,
+      "step": 523500
+    },
+    {
+      "epoch": 4.91,
+      "eval_loss": 1.010750412940979,
+      "eval_runtime": 221.9772,
+      "eval_samples_per_second": 450.497,
+      "eval_steps_per_second": 14.078,
+      "step": 523500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 8.26748747208198e-07,
+      "loss": 1.1772,
+      "step": 524000
+    },
+    {
+      "epoch": 4.92,
+      "eval_loss": 1.0138213634490967,
+      "eval_runtime": 221.8206,
+      "eval_samples_per_second": 450.815,
+      "eval_steps_per_second": 14.088,
+      "step": 524000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.798277059364501e-07,
+      "loss": 1.1817,
+      "step": 524500
+    },
+    {
+      "epoch": 4.92,
+      "eval_loss": 1.0085700750350952,
+      "eval_runtime": 221.8601,
+      "eval_samples_per_second": 450.735,
+      "eval_steps_per_second": 14.085,
+      "step": 524500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.329066646647023e-07,
+      "loss": 1.1755,
+      "step": 525000
+    },
+    {
+      "epoch": 4.93,
+      "eval_loss": 1.0190365314483643,
+      "eval_runtime": 221.8179,
+      "eval_samples_per_second": 450.82,
+      "eval_steps_per_second": 14.088,
+      "step": 525000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 6.859856233929544e-07,
+      "loss": 1.1866,
+      "step": 525500
+    },
+    {
+      "epoch": 4.93,
+      "eval_loss": 1.0140148401260376,
+      "eval_runtime": 221.8188,
+      "eval_samples_per_second": 450.818,
+      "eval_steps_per_second": 14.088,
+      "step": 525500
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 6.390645821212064e-07,
+      "loss": 1.1818,
+      "step": 526000
+    },
+    {
+      "epoch": 4.94,
+      "eval_loss": 1.0161159038543701,
+      "eval_runtime": 221.8238,
+      "eval_samples_per_second": 450.808,
+      "eval_steps_per_second": 14.088,
+      "step": 526000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 5.921435408494586e-07,
+      "loss": 1.1818,
+      "step": 526500
+    },
+    {
+      "epoch": 4.94,
+      "eval_loss": 1.0107953548431396,
+      "eval_runtime": 221.8183,
+      "eval_samples_per_second": 450.819,
+      "eval_steps_per_second": 14.088,
+      "step": 526500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 5.452224995777106e-07,
+      "loss": 1.1613,
+      "step": 527000
+    },
+    {
+      "epoch": 4.95,
+      "eval_loss": 1.0149364471435547,
+      "eval_runtime": 221.7834,
+      "eval_samples_per_second": 450.89,
+      "eval_steps_per_second": 14.09,
+      "step": 527000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 4.983014583059628e-07,
+      "loss": 1.1645,
+      "step": 527500
+    },
+    {
+      "epoch": 4.95,
+      "eval_loss": 1.0124648809432983,
+      "eval_runtime": 221.8169,
+      "eval_samples_per_second": 450.822,
+      "eval_steps_per_second": 14.088,
+      "step": 527500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 4.513804170342148e-07,
+      "loss": 1.193,
+      "step": 528000
+    },
+    {
+      "epoch": 4.95,
+      "eval_loss": 1.009380578994751,
+      "eval_runtime": 221.7526,
+      "eval_samples_per_second": 450.953,
+      "eval_steps_per_second": 14.092,
+      "step": 528000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 4.04459375762467e-07,
+      "loss": 1.1537,
+      "step": 528500
+    },
+    {
+      "epoch": 4.96,
+      "eval_loss": 1.0124224424362183,
+      "eval_runtime": 221.8595,
+      "eval_samples_per_second": 450.736,
+      "eval_steps_per_second": 14.085,
+      "step": 528500
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.57538334490719e-07,
+      "loss": 1.1691,
+      "step": 529000
+    },
+    {
+      "epoch": 4.96,
+      "eval_loss": 1.0114690065383911,
+      "eval_runtime": 222.0421,
+      "eval_samples_per_second": 450.365,
+      "eval_steps_per_second": 14.074,
+      "step": 529000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 3.1061729321897113e-07,
+      "loss": 1.1643,
+      "step": 529500
+    },
+    {
+      "epoch": 4.97,
+      "eval_loss": 1.0129092931747437,
+      "eval_runtime": 227.2564,
+      "eval_samples_per_second": 440.032,
+      "eval_steps_per_second": 13.751,
+      "step": 529500
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 2.6369625194722325e-07,
+      "loss": 1.1592,
+      "step": 530000
+    },
+    {
+      "epoch": 4.97,
+      "eval_loss": 1.0124961137771606,
+      "eval_runtime": 228.0531,
+      "eval_samples_per_second": 438.494,
+      "eval_steps_per_second": 13.703,
+      "step": 530000
+    }
+  ],
+  "max_steps": 532810,
+  "num_train_epochs": 5,
+  "total_flos": 5.5361381292640083e+17,
+  "trial_name": null,
+  "trial_params": null
+}