Update README.md
Browse files
README.md
CHANGED
@@ -10,6 +10,31 @@ tags:
|
|
10 |
## Model Name: Obsidian-3B-V0.5
|
11 |
|
12 |
Obsidian is a brand new series of Multimodalities Language Models.
|
13 |
-
Obsidian-3B-V0.5 is a LLaVA's version based on Capybara-3B-V1.9 which was built on top of StableLM-3B-4e1t.
|
14 |
|
15 |
-
Current finetuning and inference code is available on our GitHub repo: [Here](https://github.com/NousResearch/Obsidian)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
## Model Name: Obsidian-3B-V0.5
|
11 |
|
12 |
Obsidian is a brand new series of Multimodalities Language Models.
|
13 |
+
Obsidian-3B-V0.5 is a LLaVA's version based on [Capybara-3B-V1.9](https://huggingface.co/NousResearch/Capybara-3B-V1.9) which was built on top of [StableLM-3B-4e1t](stabilityai/stablelm-3b-4e1t). Capybara-3B-V1.9 achieves state-of-the-art performance when compared to model with similar size, even beats some 7B models.
|
14 |
|
15 |
+
Current finetuning and inference code is available on our GitHub repo: [Here](https://github.com/NousResearch/Obsidian)
|
16 |
+
|
17 |
+
## Acknowledgement
|
18 |
+
|
19 |
+
Obsidian-3B-V0.5 was developed and finetuned by Nous Research. in collaboration with Virtual Interactive.
|
20 |
+
|
21 |
+
## Model Training
|
22 |
+
|
23 |
+
Obsidian-3B-V0.5 followed the same training procedure as LLaVA 1.5
|
24 |
+
|
25 |
+
## Prompt Format
|
26 |
+
|
27 |
+
The model followed ChatML format. However, with `###` as the seperator
|
28 |
+
|
29 |
+
```
|
30 |
+
<|im_start|>user
|
31 |
+
What is this sign about?\n<image>
|
32 |
+
###
|
33 |
+
<|im_start|>assistant
|
34 |
+
The sign is about bullying, and it is placed on a black background with a red background.
|
35 |
+
###
|
36 |
+
```
|
37 |
+
|
38 |
+
## Benchmarks
|
39 |
+
|
40 |
+
Coming Soon!
|