Instructions to use Sabomako/Qwen3-Omni-30B-A3B-Thinking-heretic with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Sabomako/Qwen3-Omni-30B-A3B-Thinking-heretic with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("Sabomako/Qwen3-Omni-30B-A3B-Thinking-heretic") model = AutoModelForMultimodalLM.from_pretrained("Sabomako/Qwen3-Omni-30B-A3B-Thinking-heretic") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -7,8 +7,17 @@ tags:
|
|
| 7 |
- multimodal
|
| 8 |
library_name: transformers
|
| 9 |
pipeline_tag: any-to-any
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
# Qwen3-Omni
|
| 13 |
|
| 14 |
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
|
|
@@ -2049,4 +2058,4 @@ If you find our paper and code useful in your research, please consider giving a
|
|
| 2049 |
}
|
| 2050 |
``` -->
|
| 2051 |
|
| 2052 |
-
<br>
|
|
|
|
| 7 |
- multimodal
|
| 8 |
library_name: transformers
|
| 9 |
pipeline_tag: any-to-any
|
| 10 |
+
base_model:
|
| 11 |
+
- Qwen/Qwen3-Omni-30B-A3B-Thinking
|
| 12 |
---
|
| 13 |
|
| 14 |
+
## Performance
|
| 15 |
+
|
| 16 |
+
| Metric | This model | Original model ([Qwen/Qwen3-Omni-30B-A3B-Thinking](https://huggingface.co/Qwen/Qwen3-Omni-30B-A3B-Thinking)) |
|
| 17 |
+
| :----- | :--------: | :---------------------------: |
|
| 18 |
+
| **KL divergence** | 0.003563 | 0 *(by definition)* |
|
| 19 |
+
| **Refusals** | 7/100 | 99/100 |
|
| 20 |
+
|
| 21 |
# Qwen3-Omni
|
| 22 |
|
| 23 |
<a href="https://chat.qwen.ai/" target="_blank" style="margin: 2px;">
|
|
|
|
| 2058 |
}
|
| 2059 |
``` -->
|
| 2060 |
|
| 2061 |
+
<br>
|