End of training

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8581762608252674,
+    "eval_loss": 0.6532529592514038,
+    "eval_runtime": 55.107,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 178.108,
+    "eval_steps_per_second": 5.571,
+    "train_loss": 0.23546070093592578,
+    "train_runtime": 27167.7288,
+    "train_samples": 392702,
+    "train_samples_per_second": 57.819,
+    "train_steps_per_second": 0.452
+}

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_accuracy": 0.8581762608252674,
+    "eval_loss": 0.6532529592514038,
+    "eval_runtime": 55.107,
+    "eval_samples": 9815,
+    "eval_samples_per_second": 178.108,
+    "eval_steps_per_second": 5.571
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 0.23546070093592578,
+    "train_runtime": 27167.7288,
+    "train_samples": 392702,
+    "train_samples_per_second": 57.819,
+    "train_steps_per_second": 0.452
+}

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": 4.797099087353325e-05,
+      "loss": 0.6196,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.593383311603651e-05,
+      "loss": 0.4869,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.389667535853977e-05,
+      "loss": 0.4576,
+      "step": 1500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.185951760104303e-05,
+      "loss": 0.4414,
+      "step": 2000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.983050847457627e-05,
+      "loss": 0.4296,
+      "step": 2500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.779335071707953e-05,
+      "loss": 0.414,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.841365257259297,
+      "eval_loss": 0.4073391854763031,
+      "eval_runtime": 55.5785,
+      "eval_samples_per_second": 176.597,
+      "eval_steps_per_second": 5.524,
+      "step": 3068
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.575619295958279e-05,
+      "loss": 0.285,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.371903520208605e-05,
+      "loss": 0.2677,
+      "step": 4000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.168187744458931e-05,
+      "loss": 0.2725,
+      "step": 4500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.964471968709257e-05,
+      "loss": 0.268,
+      "step": 5000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.7607561929595827e-05,
+      "loss": 0.2697,
+      "step": 5500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.557040417209909e-05,
+      "loss": 0.269,
+      "step": 6000
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8579724910850739,
+      "eval_loss": 0.4016716480255127,
+      "eval_runtime": 55.5968,
+      "eval_samples_per_second": 176.539,
+      "eval_steps_per_second": 5.522,
+      "step": 6136
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.3533246414602346e-05,
+      "loss": 0.171,
+      "step": 6500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.15001629726206e-05,
+      "loss": 0.133,
+      "step": 7000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.946300521512386e-05,
+      "loss": 0.141,
+      "step": 7500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.742584745762712e-05,
+      "loss": 0.1342,
+      "step": 8000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.538868970013038e-05,
+      "loss": 0.1371,
+      "step": 8500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.3351531942633638e-05,
+      "loss": 0.1355,
+      "step": 9000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8556291390728477,
+      "eval_loss": 0.5104126930236816,
+      "eval_runtime": 55.6164,
+      "eval_samples_per_second": 176.477,
+      "eval_steps_per_second": 5.52,
+      "step": 9204
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.1314374185136898e-05,
+      "loss": 0.091,
+      "step": 9500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.277216427640156e-06,
+      "loss": 0.0689,
+      "step": 10000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.240058670143417e-06,
+      "loss": 0.0651,
+      "step": 10500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 5.2029009126466755e-06,
+      "loss": 0.0624,
+      "step": 11000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 3.1698174706649287e-06,
+      "loss": 0.0643,
+      "step": 11500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.1367340286831812e-06,
+      "loss": 0.0636,
+      "step": 12000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8581762608252674,
+      "eval_loss": 0.6532529592514038,
+      "eval_runtime": 55.631,
+      "eval_samples_per_second": 176.43,
+      "eval_steps_per_second": 5.519,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "step": 12272,
+      "total_flos": 1.4638901871267594e+18,
+      "train_loss": 0.23546070093592578,
+      "train_runtime": 27167.7288,
+      "train_samples_per_second": 57.819,
+      "train_steps_per_second": 0.452
+    }
+  ],
+  "max_steps": 12272,
+  "num_train_epochs": 4,
+  "total_flos": 1.4638901871267594e+18,
+  "trial_name": null,
+  "trial_params": null
+}