Release 1st verison
Browse files
README.md
CHANGED
|
@@ -7,15 +7,8 @@ tags:
|
|
| 7 |
- KoRWKV
|
| 8 |
---
|
| 9 |
|
| 10 |
-
> 🚧 Note: this repo is under construction, current uploaded version is >140 Billion tokens trained ckpt 🚧
|
| 11 |
-
>
|
| 12 |
> Instruction-Finetuned model is available at [beomi/KoAlpaca-KoRWKV-6B](https://huggingface.co/beomi/KoAlpaca-KoRWKV-6B)
|
| 13 |
|
| 14 |
-
## Todo
|
| 15 |
-
|
| 16 |
-
- ⏳ Train 6B
|
| 17 |
-
|
| 18 |
-
|
| 19 |
# KoRWKV Model Card
|
| 20 |
|
| 21 |
KoRWKV (6B) trained on Korean dataset with RWKVv4 Neo Architecture.
|
|
@@ -28,11 +21,11 @@ Junbum Lee (aka Beomi)
|
|
| 28 |
|
| 29 |
**Model date**
|
| 30 |
|
| 31 |
-
KoRWKV was trained between
|
| 32 |
|
| 33 |
**Model version**
|
| 34 |
|
| 35 |
-
This is
|
| 36 |
|
| 37 |
**Model type**
|
| 38 |
|
|
|
|
| 7 |
- KoRWKV
|
| 8 |
---
|
| 9 |
|
|
|
|
|
|
|
| 10 |
> Instruction-Finetuned model is available at [beomi/KoAlpaca-KoRWKV-6B](https://huggingface.co/beomi/KoAlpaca-KoRWKV-6B)
|
| 11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
# KoRWKV Model Card
|
| 13 |
|
| 14 |
KoRWKV (6B) trained on Korean dataset with RWKVv4 Neo Architecture.
|
|
|
|
| 21 |
|
| 22 |
**Model date**
|
| 23 |
|
| 24 |
+
KoRWKV was trained between 2023.05~2023.07
|
| 25 |
|
| 26 |
**Model version**
|
| 27 |
|
| 28 |
+
This is 1st release of the model.
|
| 29 |
|
| 30 |
**Model type**
|
| 31 |
|