Update README.md
Browse files
README.md
CHANGED
|
@@ -30,6 +30,7 @@ It achieves the following results on the evaluation set:
|
|
| 30 |
- Average token count: 15.45
|
| 31 |
|
| 32 |
## Model description
|
|
|
|
| 33 |
|
| 34 |
Loss is a combination of:
|
| 35 |
- CrossEntropy
|
|
@@ -38,7 +39,7 @@ Loss is a combination of:
|
|
| 38 |
Loss = theta * Custom loss + (1 - theta) * CrossEntropy
|
| 39 |
|
| 40 |
(theta = 0.3)
|
| 41 |
-
|
| 42 |
## Intended uses & limitations
|
| 43 |
|
| 44 |
More information needed
|
|
|
|
| 30 |
- Average token count: 15.45
|
| 31 |
|
| 32 |
## Model description
|
| 33 |
+
Data is cleaned and preprocessed: "summarize" prefix added for each original text input.
|
| 34 |
|
| 35 |
Loss is a combination of:
|
| 36 |
- CrossEntropy
|
|
|
|
| 39 |
Loss = theta * Custom loss + (1 - theta) * CrossEntropy
|
| 40 |
|
| 41 |
(theta = 0.3)
|
| 42 |
+
|
| 43 |
## Intended uses & limitations
|
| 44 |
|
| 45 |
More information needed
|