End of training

README.md

@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [cardiffnlp/twitter-roberta-base-2019-90m](https://huggingface.co/cardiffnlp/twitter-roberta-base-2019-90m) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.9349
+- Loss: 1.9422
 
 ## Model description
 

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 2.62,
-    "eval_loss": 2.006091594696045,
-    "eval_runtime": 861.4578,
-    "eval_samples": 771820,
-    "eval_samples_per_second": 895.946,
-    "eval_steps_per_second": 55.997,
-    "perplexity": 7.4342046074668735,
-    "train_loss": 2.133689431966146,
-    "train_runtime": 416842.919,
-    "train_samples": 14664588,
-    "train_samples_per_second": 92.121,
-    "train_steps_per_second": 5.758
+    "epoch": 5.05,
+    "eval_loss": 1.942165732383728,
+    "eval_runtime": 459.5033,
+    "eval_samples": 400000,
+    "eval_samples_per_second": 870.505,
+    "eval_steps_per_second": 54.407,
+    "perplexity": 6.97383809440389,
+    "train_loss": 2.0860133251953124,
+    "train_runtime": 289964.3288,
+    "train_samples": 7600000,
+    "train_samples_per_second": 132.43,
+    "train_steps_per_second": 8.277
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.62,
-    "eval_loss": 2.006091594696045,
-    "eval_runtime": 861.4578,
-    "eval_samples": 771820,
-    "eval_samples_per_second": 895.946,
-    "eval_steps_per_second": 55.997,
-    "perplexity": 7.4342046074668735
+    "epoch": 5.05,
+    "eval_loss": 1.942165732383728,
+    "eval_runtime": 459.5033,
+    "eval_samples": 400000,
+    "eval_samples_per_second": 870.505,
+    "eval_steps_per_second": 54.407,
+    "perplexity": 6.97383809440389
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.62,
-    "train_loss": 2.133689431966146,
-    "train_runtime": 416842.919,
-    "train_samples": 14664588,
-    "train_samples_per_second": 92.121,
-    "train_steps_per_second": 5.758
+    "epoch": 5.05,
+    "train_loss": 2.0860133251953124,
+    "train_runtime": 289964.3288,
+    "train_samples": 7600000,
+    "train_samples_per_second": 132.43,
+    "train_steps_per_second": 8.277
 }