Update README.md
Browse files
README.md
CHANGED
|
@@ -31,7 +31,7 @@ Cephalo provides a robust framework for multimodal interaction and understanding
|
|
| 31 |
- [Cephalo-Llava-v1.6-Mistral-8b-alpha](https://huggingface.co/lamm-mit/Cephalo-Llava-v1.6-Mistral-8b-alpha)
|
| 32 |
- Trained on GPT-4o distilled image-text data from Wikipedia, with low-resolution images. Does not perform well on multiple image queries, and has some inconsistencies in understanding.
|
| 33 |
|
| 34 |
-
### Merged models
|
| 35 |
|
| 36 |
- [Cephalo-Idefics-2-vision-10b-alpha](https://huggingface.co/lamm-mit/Cephalo-Idefics-2-vision-10b-alpha)
|
| 37 |
- Merged model, 32+8=40 layers, checkpoint after first epoch. Trained on GPT-4o distilled image-text data from Wikipedia and scientific papers.
|
|
|
|
| 31 |
- [Cephalo-Llava-v1.6-Mistral-8b-alpha](https://huggingface.co/lamm-mit/Cephalo-Llava-v1.6-Mistral-8b-alpha)
|
| 32 |
- Trained on GPT-4o distilled image-text data from Wikipedia, with low-resolution images. Does not perform well on multiple image queries, and has some inconsistencies in understanding.
|
| 33 |
|
| 34 |
+
### Merged 10b models
|
| 35 |
|
| 36 |
- [Cephalo-Idefics-2-vision-10b-alpha](https://huggingface.co/lamm-mit/Cephalo-Idefics-2-vision-10b-alpha)
|
| 37 |
- Merged model, 32+8=40 layers, checkpoint after first epoch. Trained on GPT-4o distilled image-text data from Wikipedia and scientific papers.
|