End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9125022881200805,
+    "eval_loss": 0.3208048343658447,
+    "eval_runtime": 24.4432,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 223.498,
+    "eval_steps_per_second": 27.942,
+    "train_loss": 0.21106615146146823,
+    "train_runtime": 3939.5621,
+    "train_samples": 104743,
+    "train_samples_per_second": 79.762,
+    "train_steps_per_second": 2.493
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 3.0,
+    "eval_accuracy": 0.9125022881200805,
+    "eval_loss": 0.3208048343658447,
+    "eval_runtime": 24.4432,
+    "eval_samples": 5463,
+    "eval_samples_per_second": 223.498,
+    "eval_steps_per_second": 27.942
+}

runs/Jun21_16-23-57_mewtwo/events.out.tfevents.1655832943.mewtwo

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88a09a5a267bba335a2c4174fa27357f56cacfb079231b82f58debdca7d117cd
+size 40

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.21106615146146823,
+    "train_runtime": 3939.5621,
+    "train_samples": 104743,
+    "train_samples_per_second": 79.762,
+    "train_steps_per_second": 2.493
+}

trainer_state.json

@@ -0,0 +1,166 @@
+{
+  "best_metric": 0.9125022881200805,
+  "best_model_checkpoint": "./fine-tune/bert-base-uncased/qnli/checkpoint-9822",
+  "epoch": 3.0,
+  "global_step": 9822,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.8981877418041133e-05,
+      "loss": 0.4366,
+      "step": 500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7963754836082265e-05,
+      "loss": 0.354,
+      "step": 1000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.69456322541234e-05,
+      "loss": 0.3286,
+      "step": 1500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.592750967216453e-05,
+      "loss": 0.3075,
+      "step": 2000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4909387090205662e-05,
+      "loss": 0.3107,
+      "step": 2500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.3891264508246794e-05,
+      "loss": 0.289,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9093904448105437,
+      "eval_loss": 0.22888238728046417,
+      "eval_runtime": 26.3583,
+      "eval_samples_per_second": 207.259,
+      "eval_steps_per_second": 25.912,
+      "step": 3274
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.2873141926287927e-05,
+      "loss": 0.2416,
+      "step": 3500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.1855019344329057e-05,
+      "loss": 0.1885,
+      "step": 4000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.083689676237019e-05,
+      "loss": 0.1864,
+      "step": 4500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.818774180411322e-06,
+      "loss": 0.1906,
+      "step": 5000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.800651598452455e-06,
+      "loss": 0.1946,
+      "step": 5500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.782529016493586e-06,
+      "loss": 0.1904,
+      "step": 6000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 6.764406434534719e-06,
+      "loss": 0.1801,
+      "step": 6500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9117700896943072,
+      "eval_loss": 0.24933215975761414,
+      "eval_runtime": 26.3984,
+      "eval_samples_per_second": 206.945,
+      "eval_steps_per_second": 25.873,
+      "step": 6548
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.74628385257585e-06,
+      "loss": 0.1167,
+      "step": 7000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.728161270616982e-06,
+      "loss": 0.1129,
+      "step": 7500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 3.7100386886581147e-06,
+      "loss": 0.1138,
+      "step": 8000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.6919161066992467e-06,
+      "loss": 0.108,
+      "step": 8500
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.6737935247403788e-06,
+      "loss": 0.1165,
+      "step": 9000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.55670942781511e-07,
+      "loss": 0.1074,
+      "step": 9500
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9125022881200805,
+      "eval_loss": 0.3208048343658447,
+      "eval_runtime": 26.8523,
+      "eval_samples_per_second": 203.446,
+      "eval_steps_per_second": 25.435,
+      "step": 9822
+    },
+    {
+      "epoch": 3.0,
+      "step": 9822,
+      "total_flos": 2.066928095367936e+16,
+      "train_loss": 0.21106615146146823,
+      "train_runtime": 3939.5621,
+      "train_samples_per_second": 79.762,
+      "train_steps_per_second": 2.493
+    }
+  ],
+  "max_steps": 9822,
+  "num_train_epochs": 3,
+  "total_flos": 2.066928095367936e+16,
+  "trial_name": null,
+  "trial_params": null
+}