Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model:
|
| 3 |
+
- ddh0/Cassiopeia-70B
|
| 4 |
+
- Sao10K/Llama-3.3-70B-Vulpecula-r1
|
| 5 |
+
base_model_relation: merge
|
| 6 |
+
license: unknown
|
| 7 |
+
thumbnail: https://huggingface.co/ddh0/Andromeda-70B/resolve/main/andromeda.png
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# Andromeda-70B
|
| 11 |
+
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
**Andromeda-70B** is the result of an experimental SLERP merge of [Cassiopeia-70B](https://huggingface.co/ddh0/Cassiopeia-70B) and [Sao10K/Llama-3.3-70B-Vulpecula-r1](https://huggingface.co/Sao10K/Llama-3.3-70B-Vulpecula-r1). It is a coherent, unaligned, general-purpose model intended to be used for creative tasks such as storywriting, brainstorming, interactive roleplay, etc.
|
| 15 |
+
|
| 16 |
+
## Merge composition
|
| 17 |
+
|
| 18 |
+
```yaml
|
| 19 |
+
models:
|
| 20 |
+
- model: /opt/workspace/hf/Cassiopeia-70B
|
| 21 |
+
- model: /opt/workspace/hf/Llama-3.3-70B-Vulpecula-r1
|
| 22 |
+
merge_method: slerp
|
| 23 |
+
base_model: /opt/workspace/hf/Cassiopeia-70B
|
| 24 |
+
parameters:
|
| 25 |
+
t: 0.7
|
| 26 |
+
dtype: bfloat16
|
| 27 |
+
```
|
| 28 |
+
|
| 29 |
+
## Feedback
|
| 30 |
+
|
| 31 |
+
If you like this model, please support [Sao10k](https://sao10k.carrd.co).
|
| 32 |
+
|
| 33 |
+
Feedback on this merge is very welcome, good or bad! Please leave a comment in this discussion with your thoughts: [Andromeda-70B/discussions/1](https://huggingface.co/ddh0/Andromeda-70B/discussions/1)
|