Update README.md
Browse files
README.md
CHANGED
@@ -7,7 +7,15 @@ tags:
|
|
7 |
- udkai/Turdus
|
8 |
- flemmingmiguel/DareBeagle-7B
|
9 |
---
|
|
|
10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
# TurdusDareBeagle-7B
|
12 |
|
13 |
TurdusDareBeagle-7B is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
|
|
7 |
- udkai/Turdus
|
8 |
- flemmingmiguel/DareBeagle-7B
|
9 |
---
|
10 |
+
## Exl2 version of [Undi95/OpenDolphinMaid-4x7b](https://huggingface.co/Undi95/OpenDolphinMaid-4x7b)
|
11 |
|
12 |
+
## branch
|
13 |
+
7bh8 : 7bpw h8
|
14 |
+
|
15 |
+
Using ThePile [0007.parquet](https://huggingface.co/datasets/EleutherAI/the_pile_deduplicated/resolve/refs%2Fconvert%2Fparquet/default/train/0007.parquet) as dataset
|
16 |
+
|
17 |
+
Quantization settings : ```python convert.py -i models/flemmingmiguel_TurdusDareBeagle-7B -o TurdusDareBeagle-7B-temp -cf TurdusDareBeagle-7B-8bpw-h8-exl2 -c 0007.parquet -l 8192 -b 8 -hb 8 -ml 8192```
|
18 |
+
### below this line is original readme
|
19 |
# TurdusDareBeagle-7B
|
20 |
|
21 |
TurdusDareBeagle-7B is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|