Update README.md
Browse files
README.md
CHANGED
|
@@ -12,6 +12,8 @@ metrics:
|
|
| 12 |
**(!) _Don't forget to preprocess unknown_tokens and substitute them with <|endoftext|>. Otherwise the \<unk\> tokens in dataset will be split into the '<', 'unk' and '>' tokens_**
|
| 13 |
|
| 14 |
|
|
|
|
|
|
|
| 15 |
**Dependence of the cross entropy loss on the length of the context for prediction**
|
| 16 |
|
| 17 |
- x-axis*128 = context length
|
|
|
|
| 12 |
**(!) _Don't forget to preprocess unknown_tokens and substitute them with <|endoftext|>. Otherwise the \<unk\> tokens in dataset will be split into the '<', 'unk' and '>' tokens_**
|
| 13 |
|
| 14 |
|
| 15 |
+
- Full context (1024) perplexity: **13.68**
|
| 16 |
+
|
| 17 |
**Dependence of the cross entropy loss on the length of the context for prediction**
|
| 18 |
|
| 19 |
- x-axis*128 = context length
|