End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8332144676515537,
+    "eval_loss": 0.6082797050476074,
+    "eval_runtime": 30.7564,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 319.121,
+    "eval_steps_per_second": 5.007,
+    "train_loss": 0.31909979493241886,
+    "train_runtime": 16147.6692,
+    "train_samples": 392702,
+    "train_samples_per_second": 97.278,
+    "train_steps_per_second": 0.76
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8332144676515537,
+    "eval_loss": 0.6082797050476074,
+    "eval_runtime": 30.7564,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 319.121,
+    "eval_steps_per_second": 5.007
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.31909979493241886,
+    "train_runtime": 16147.6692,
+    "train_samples": 392702,
+    "train_samples_per_second": 97.278,
+    "train_steps_per_second": 0.76
+}

trainer_state.json

@@ -0,0 +1,205 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "global_step": 12272,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7962842242503265e-05,
+      "loss": 0.7079,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.592568448500652e-05,
+      "loss": 0.5696,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3888526727509784e-05,
+      "loss": 0.5353,
+      "step": 1500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.185136897001304e-05,
+      "loss": 0.5154,
+      "step": 2000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.98142112125163e-05,
+      "loss": 0.4966,
+      "step": 2500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.777705345501956e-05,
+      "loss": 0.4804,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8202750891492613,
+      "eval_loss": 0.46233224868774414,
+      "eval_runtime": 31.0795,
+      "eval_samples_per_second": 315.803,
+      "eval_steps_per_second": 4.955,
+      "step": 3068
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.5739895697522816e-05,
+      "loss": 0.3742,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.370273794002607e-05,
+      "loss": 0.3565,
+      "step": 4000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.1669654498044335e-05,
+      "loss": 0.3614,
+      "step": 4500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.9632496740547587e-05,
+      "loss": 0.3558,
+      "step": 5000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.759533898305085e-05,
+      "loss": 0.3589,
+      "step": 5500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.5558181225554106e-05,
+      "loss": 0.3583,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8309730005094244,
+      "eval_loss": 0.4523915946483612,
+      "eval_runtime": 30.8881,
+      "eval_samples_per_second": 317.76,
+      "eval_steps_per_second": 4.986,
+      "step": 6136
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3525097783572363e-05,
+      "loss": 0.2631,
+      "step": 6500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.148794002607562e-05,
+      "loss": 0.2303,
+      "step": 7000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9450782268578878e-05,
+      "loss": 0.2311,
+      "step": 7500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.741769882659713e-05,
+      "loss": 0.2248,
+      "step": 8000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.5380541069100394e-05,
+      "loss": 0.2302,
+      "step": 8500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3343383311603652e-05,
+      "loss": 0.2257,
+      "step": 9000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8346408558329088,
+      "eval_loss": 0.5004044771194458,
+      "eval_runtime": 30.8675,
+      "eval_samples_per_second": 317.972,
+      "eval_steps_per_second": 4.989,
+      "step": 9204
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.1306225554106911e-05,
+      "loss": 0.1736,
+      "step": 9500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.26906779661017e-06,
+      "loss": 0.145,
+      "step": 10000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.235984354628423e-06,
+      "loss": 0.1434,
+      "step": 10500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.198826597131682e-06,
+      "loss": 0.1398,
+      "step": 11000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1616688396349417e-06,
+      "loss": 0.141,
+      "step": 11500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1245110821382008e-06,
+      "loss": 0.1402,
+      "step": 12000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8332144676515537,
+      "eval_loss": 0.6082797050476074,
+      "eval_runtime": 30.8609,
+      "eval_samples_per_second": 318.04,
+      "eval_steps_per_second": 4.99,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "step": 12272,
+      "total_flos": 4.133006614744842e+17,
+      "train_loss": 0.31909979493241886,
+      "train_runtime": 16147.6692,
+      "train_samples_per_second": 97.278,
+      "train_steps_per_second": 0.76
+    }
+  ],
+  "max_steps": 12272,
+  "num_train_epochs": 4,
+  "total_flos": 4.133006614744842e+17,
+  "trial_name": null,
+  "trial_params": null
+}