traintogpb
commited on
Commit
•
1faf122
1
Parent(s):
1163aed
Update README.md
Browse files
README.md
CHANGED
@@ -35,7 +35,7 @@ pipeline_tag: translation
|
|
35 |
- Merge adapters and upscaled in BrainFloat 16-bit precision
|
36 |
|
37 |
### Usage (IMPORTANT)
|
38 |
-
- Should remove the EOS token
|
39 |
```python
|
40 |
# MODEL
|
41 |
model_name = 'beomi/Llama-3-Open-Ko-8B'
|
@@ -80,4 +80,4 @@ pipeline_tag: translation
|
|
80 |
### Framework versions
|
81 |
|
82 |
- PEFT 0.8.2
|
83 |
-
|
|
|
35 |
- Merge adapters and upscaled in BrainFloat 16-bit precision
|
36 |
|
37 |
### Usage (IMPORTANT)
|
38 |
+
- Should remove the EOS token at the end of the prompt.
|
39 |
```python
|
40 |
# MODEL
|
41 |
model_name = 'beomi/Llama-3-Open-Ko-8B'
|
|
|
80 |
### Framework versions
|
81 |
|
82 |
- PEFT 0.8.2
|
83 |
+
|