diff --git a/trainer_state.json b/trainer_state.json
new file mode 100644
index 0000000..f26431b
--- /dev/null
+++ b/trainer_state.json
@@ -0,0 +1,70 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.991735537190083,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.66,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.3311,
+      "step": 100
+    },
+    {
+      "epoch": 1.66,
+      "eval_accuracy": 0.6033057570457458,
+      "eval_loss": 1.0739655494689941,
+      "eval_runtime": 74.6016,
+      "eval_samples_per_second": 1.622,
+      "eval_steps_per_second": 0.416,
+      "step": 100
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.6235,
+      "step": 200
+    },
+    {
+      "epoch": 3.33,
+      "eval_accuracy": 0.8512396812438965,
+      "eval_loss": 0.47899967432022095,
+      "eval_runtime": 71.6672,
+      "eval_samples_per_second": 1.688,
+      "eval_steps_per_second": 0.433,
+      "step": 200
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.0,
+      "loss": 0.2154,
+      "step": 300
+    },
+    {
+      "epoch": 4.99,
+      "eval_accuracy": 0.9008264541625977,
+      "eval_loss": 0.3769935369491577,
+      "eval_runtime": 73.3118,
+      "eval_samples_per_second": 1.65,
+      "eval_steps_per_second": 0.423,
+      "step": 300
+    },
+    {
+      "epoch": 4.99,
+      "step": 300,
+      "total_flos": 4.279570856908188e+17,
+      "train_loss": 0.7233316548665365,
+      "train_runtime": 4793.1545,
+      "train_samples_per_second": 0.504,
+      "train_steps_per_second": 0.063
+    }
+  ],
+  "max_steps": 300,
+  "num_train_epochs": 5,
+  "total_flos": 4.279570856908188e+17,
+  "trial_name": null,
+  "trial_params": null
+}