Commit
•
2037622
1
Parent(s):
eb30cc5
Update README.md
Browse files
README.md
CHANGED
|
@@ -19,7 +19,7 @@ This model is a fine tuned version of the facebook/nllb-200-disillted-600M model
|
|
| 19 |
|
| 20 |
The model was fit with a custom tokenizer and language tag for Amis and additional Traditional Chinese tokens.
|
| 21 |
|
| 22 |
-
The model was trained on roughly ~90k sentence pairs for
|
| 23 |
|
| 24 |
## Evaluation
|
| 25 |
|
|
|
|
| 19 |
|
| 20 |
The model was fit with a custom tokenizer and language tag for Amis and additional Traditional Chinese tokens.
|
| 21 |
|
| 22 |
+
The model was trained on roughly ~90k sentence pairs for 60k Epochs until validation loss plateaued
|
| 23 |
|
| 24 |
## Evaluation
|
| 25 |
|