Commit
•
f4b2ca8
1
Parent(s):
ee42f5c
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: other
|
| 3 |
license_name: yi-license
|
|
|
|
| 1 |
+
NousResearch/Nous-Capybara-34B and migtissera/Tess-M-Creative-v1.0 ties merged with mergekit, using the following config:
|
| 2 |
+
|
| 3 |
+
```
|
| 4 |
+
models:
|
| 5 |
+
- model: /home/alpha/Storage/Models/Raw/larryvrh_Yi-34B-200K-Llamafied
|
| 6 |
+
# no parameters necessary for base model
|
| 7 |
+
- model: /home/alpha/Storage/Models/Raw/migtissera_Tess-M-v1.0
|
| 8 |
+
parameters:
|
| 9 |
+
density: 0.6
|
| 10 |
+
weight: 1.0
|
| 11 |
+
- model: /home/alpha/Storage/Models/Raw/Nous-Capybara-34B
|
| 12 |
+
parameters:
|
| 13 |
+
density: 0.6
|
| 14 |
+
weight: 1.0
|
| 15 |
+
merge_method: ties
|
| 16 |
+
base_model: //home/alpha/Storage/Models/Raw/larryvrh_Yi-34B-200K-Llamafied
|
| 17 |
+
parameters:
|
| 18 |
+
normalize: true
|
| 19 |
+
int8_mask: true
|
| 20 |
+
dtype: float16
|
| 21 |
+
```
|
| 22 |
+
|
| 23 |
+
Both are 200K context models with Vicuna syntax, so:
|
| 24 |
+
|
| 25 |
+
# Prompt Format:
|
| 26 |
+
|
| 27 |
+
```
|
| 28 |
+
SYSTEM: <ANY SYSTEM CONTEXT>
|
| 29 |
+
USER: <USER QUERY>
|
| 30 |
+
ASSISTANT: ...
|
| 31 |
+
```
|
| 32 |
+
Stop token: `</s>`
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
Credits:
|
| 36 |
+
https://github.com/cg123/mergekit
|
| 37 |
+
https://huggingface.co/NousResearch/Nous-Capybara-34B/discussions
|
| 38 |
+
https://huggingface.co/migtissera/Tess-M-Creative-v1.0
|
| 39 |
+
https://huggingface.co/larryvrh/Yi-34B-200K-Llamafied
|
| 40 |
+
https://huggingface.co/01-ai/Yi-34B-200K
|
| 41 |
+
|
| 42 |
+
|
| 43 |
---
|
| 44 |
license: other
|
| 45 |
license_name: yi-license
|