Update README.md
Browse files
README.md
CHANGED
|
@@ -36,6 +36,11 @@ Llama-2-Ko is an auto-regressive language model that uses an optimized transform
|
|
| 36 |
|---|---|---|---|---|---|---|
|
| 37 |
|Llama 2|*A new mix of Korean online data*|7B|4k|✗|160B|3.0 x 10<sup>-4</sup>|
|
| 38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
---
|
| 41 |
|
|
|
|
| 36 |
|---|---|---|---|---|---|---|
|
| 37 |
|Llama 2|*A new mix of Korean online data*|7B|4k|✗|160B|3.0 x 10<sup>-4</sup>|
|
| 38 |
|
| 39 |
+
**Vocab Expansion**
|
| 40 |
+
|
| 41 |
+
- Original Llama-2: 32000 Sentencepiece BPE
|
| 42 |
+
- **Expanded Llama-2-ko: 46336** Sentencepiece BPE
|
| 43 |
+
- New vocab and merges, trained with Korean Corpus
|
| 44 |
|
| 45 |
---
|
| 46 |
|