Update README.md about ComfyUI
Browse files
README.md
CHANGED
|
@@ -39,10 +39,11 @@ The framework leverages advanced diffusion models and transformer architectures
|
|
| 39 |
|
| 40 |
## 🔥 News
|
| 41 |
|
| 42 |
-
* **[2026.02.
|
|
|
|
| 43 |
|
| 44 |
## 📝 TODO List
|
| 45 |
-
- [
|
| 46 |
- [ ] Release our unified creation model.
|
| 47 |
- [ ] Release training code.
|
| 48 |
|
|
@@ -158,7 +159,7 @@ python inference.py \
|
|
| 158 |
python inference.py \
|
| 159 |
--pretrained_model_name_or_path ./ckpts \
|
| 160 |
--media_path ./assets/examples/video1.mp4 \
|
| 161 |
-
--prompt "Replace the monkey with Ultraman.
|
| 162 |
--output_path ./results/test_single_output/tv2v \
|
| 163 |
--num_inference_steps 50 \
|
| 164 |
--num_frames 81 \
|
|
@@ -251,6 +252,24 @@ accelerate launch --config_file acc_config/accelerate_config.yaml --num_processe
|
|
| 251 |
--rewrite_instruction
|
| 252 |
```
|
| 253 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 254 |
## ⚙️ Configuration Details
|
| 255 |
|
| 256 |
### Task Types
|
|
|
|
| 39 |
|
| 40 |
## 🔥 News
|
| 41 |
|
| 42 |
+
* **[2026.02.20]** 🎨 Added [ComfyUI support](#-comfyui-support) with custom nodes for all task types (T2I, T2V, TI2I, TV2V).
|
| 43 |
+
* **[2026.02.17]** 🚀 Initial release v0.1 of the Capybara inference framework supporting generation and instruction-based editing tasks (T2I, T2V, TI2I, TV2V).
|
| 44 |
|
| 45 |
## 📝 TODO List
|
| 46 |
+
- [x] Add support for ComfyUI.
|
| 47 |
- [ ] Release our unified creation model.
|
| 48 |
- [ ] Release training code.
|
| 49 |
|
|
|
|
| 159 |
python inference.py \
|
| 160 |
--pretrained_model_name_or_path ./ckpts \
|
| 161 |
--media_path ./assets/examples/video1.mp4 \
|
| 162 |
+
--prompt "Replace the monkey with Ultraman. Keep the Ultraman's motion matched the original running pose and motion of monkey." \
|
| 163 |
--output_path ./results/test_single_output/tv2v \
|
| 164 |
--num_inference_steps 50 \
|
| 165 |
--num_frames 81 \
|
|
|
|
| 252 |
--rewrite_instruction
|
| 253 |
```
|
| 254 |
|
| 255 |
+
## 🎨 ComfyUI Support
|
| 256 |
+
|
| 257 |
+
Capybara provides custom ComfyUI nodes for all task types (T2V, T2I, TI2I, TV2V).
|
| 258 |
+
|
| 259 |
+
```bash
|
| 260 |
+
ln -s /path/to/Capybara /path/to/ComfyUI/custom_nodes/Capybara
|
| 261 |
+
```
|
| 262 |
+
|
| 263 |
+
| Node | Description |
|
| 264 |
+
| --- | --- |
|
| 265 |
+
| **Capybara Load Pipeline** | Load all model components with automatic attention backend selection |
|
| 266 |
+
| **Capybara Generate** | Main generation / editing node for all task types |
|
| 267 |
+
| **Capybara Load Video** | Load a video file as IMAGE frames + fps |
|
| 268 |
+
| **Capybara Load Rewrite Model** | Load Qwen3-VL for prompt rewriting |
|
| 269 |
+
| **Capybara Rewrite Instruction** | Expand short prompts into detailed instructions |
|
| 270 |
+
|
| 271 |
+
A sample workflow is provided in [`comfyui/examples/`](comfyui/examples/). For setup details and node documentation, see the [ComfyUI README](comfyui/README.md).
|
| 272 |
+
|
| 273 |
## ⚙️ Configuration Details
|
| 274 |
|
| 275 |
### Task Types
|