Update README.md
Browse files
README.md
CHANGED
|
@@ -38,6 +38,11 @@ Cephalo provides a robust framework for multimodal interaction and understanding
|
|
| 38 |
- [Cephalo-Idefics-2-vision-10b-beta](https://huggingface.co/lamm-mit/Cephalo-Idefics-2-vision-10b-beta)
|
| 39 |
- Merged model, 32+8=40 layers, checkpoint after second epoch. Trained on GPT-4o distilled image-text data from Wikipedia and scientific papers.
|
| 40 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
### Additional codes and tools
|
| 42 |
|
| 43 |
Additional codes and tools are provided at [https://github.com/lamm-mit/Cephalo](https://github.com/lamm-mit/Cephalo).
|
|
|
|
| 38 |
- [Cephalo-Idefics-2-vision-10b-beta](https://huggingface.co/lamm-mit/Cephalo-Idefics-2-vision-10b-beta)
|
| 39 |
- Merged model, 32+8=40 layers, checkpoint after second epoch. Trained on GPT-4o distilled image-text data from Wikipedia and scientific papers.
|
| 40 |
|
| 41 |
+
### Merged 12b models
|
| 42 |
+
|
| 43 |
+
- [lamm-mit/Cephalo-Idefics-2-vision-12b-alpha](https://huggingface.co/lamm-mit/Cephalo-Idefics-2-vision-12b-alpha)
|
| 44 |
+
- Merged model, 32+16=48 layers, checkpoint after first epoch. Trained on GPT-4o distilled image-text data from Wikipedia and scientific papers (dataset derivived from both Idefics-2 and GPT-4o distillation of the paper corpus).
|
| 45 |
+
|
| 46 |
### Additional codes and tools
|
| 47 |
|
| 48 |
Additional codes and tools are provided at [https://github.com/lamm-mit/Cephalo](https://github.com/lamm-mit/Cephalo).
|