umn-cyber/indobert-hoax-detection

README.md

@@ -21,11 +21,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [indobenchmark/indobert-base-p1](https://huggingface.co/indobenchmark/indobert-base-p1) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0480
-- Accuracy: 0.9885
-- F1: 0.9879
-- Precision: 0.9879
-- Recall: 0.9879
+- Loss: 0.0556
+- Accuracy: 0.9831
+- F1: 0.9823
+- Precision: 0.9781
+- Recall: 0.9865
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9847767253044655,
-    "eval_f1": 0.9839686498040613,
-    "eval_loss": 0.05431538447737694,
-    "eval_precision": 0.9857244825124911,
-    "eval_recall": 0.9822190611664295,
-    "eval_runtime": 82.271,
-    "eval_samples_per_second": 35.93,
-    "eval_steps_per_second": 1.13,
-    "total_flos": 3.11005644392448e+16,
-    "train_loss": 0.03474135211743263,
-    "train_runtime": 19879.8681,
-    "train_samples_per_second": 5.946,
-    "train_steps_per_second": 0.186
+    "epoch": 3.0,
+    "eval_accuracy": 0.983085250338295,
+    "eval_f1": 0.9822946175637394,
+    "eval_loss": 0.055622417479753494,
+    "eval_precision": 0.9781382228490832,
+    "eval_recall": 0.9864864864864865,
+    "eval_runtime": 83.5515,
+    "eval_samples_per_second": 35.379,
+    "eval_steps_per_second": 1.113,
+    "total_flos": 1.86598360461312e+16,
+    "train_loss": 0.0482267698942674,
+    "train_runtime": 11974.8731,
+    "train_samples_per_second": 5.922,
+    "train_steps_per_second": 0.185
 }

eval_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9861252115059221,
-    "eval_f1": 0.9854144432586268,
-    "eval_loss": 0.049318719655275345,
-    "eval_precision": 0.9857651245551602,
-    "eval_recall": 0.9850640113798008,
-    "eval_runtime": 82.1767,
-    "eval_samples_per_second": 35.959,
-    "eval_steps_per_second": 1.132
+    "epoch": 3.0,
+    "eval_accuracy": 0.9868020304568528,
+    "eval_f1": 0.9861946902654867,
+    "eval_loss": 0.04358154907822609,
+    "eval_precision": 0.981677237491191,
+    "eval_recall": 0.9907539118065434,
+    "eval_runtime": 83.335,
+    "eval_samples_per_second": 35.459,
+    "eval_steps_per_second": 1.116
 }

test_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 5.0,
-    "eval_accuracy": 0.9847767253044655,
-    "eval_f1": 0.9839686498040613,
-    "eval_loss": 0.05431538447737694,
-    "eval_precision": 0.9857244825124911,
-    "eval_recall": 0.9822190611664295,
-    "eval_runtime": 82.271,
-    "eval_samples_per_second": 35.93,
-    "eval_steps_per_second": 1.13
+    "epoch": 3.0,
+    "eval_accuracy": 0.983085250338295,
+    "eval_f1": 0.9822946175637394,
+    "eval_loss": 0.055622417479753494,
+    "eval_precision": 0.9781382228490832,
+    "eval_recall": 0.9864864864864865,
+    "eval_runtime": 83.5515,
+    "eval_samples_per_second": 35.379,
+    "eval_steps_per_second": 1.113
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 5.0,
-    "total_flos": 3.11005644392448e+16,
-    "train_loss": 0.03474135211743263,
-    "train_runtime": 19879.8681,
-    "train_samples_per_second": 5.946,
-    "train_steps_per_second": 0.186
+    "epoch": 3.0,
+    "total_flos": 1.86598360461312e+16,
+    "train_loss": 0.0482267698942674,
+    "train_runtime": 11974.8731,
+    "train_samples_per_second": 5.922,
+    "train_steps_per_second": 0.185
 }