Not working

### Custom Node Testing

- [ ] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)

### Expected Behavior

❌ [Qwen3-TTS] Model class is not loaded because the 'qwen_tts' package failed to import. Please check the ComfyUI console for the detailed 'Critical Import Error' above.

### Actual Behavior

❌ [Qwen3-TTS] Model class is not loaded because the 'qwen_tts' package failed to import. Please check the ComfyUI console for the detailed 'Critical Import Error' above.

### Steps to Reproduce

❌ [Qwen3-TTS] Model class is not loaded because the 'qwen_tts' package failed to import. Please check the ComfyUI console for the detailed 'Critical Import Error' above.

### Debug Logs

```powershell
# ComfyUI Error Report
## Error Details
- **Node ID:** 39
- **Node Type:** FB_Qwen3TTSCustomVoice
- **Exception Type:** RuntimeError
- **Exception Message:** ❌ [Qwen3-TTS] Model class is not loaded because the 'qwen_tts' package failed to import. Please check the ComfyUI console for the detailed 'Critical Import Error' above.

## Stack Trace

File "C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 518, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 329, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 303, in _async_map_node_over_list
await process_inputs(input_dict, i)

File "C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 291, in process_inputs
result = f(**inputs)
^^^^^^^^^^^

File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\nodes.py", line 790, in generate
model = load_qwen_model("CustomVoice", model_choice, device, precision, attention, unload_model_after_generate, previous_attention)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\nodes.py", line 374, in load_qwen_model
raise RuntimeError(

## System Information
- **ComfyUI Version:** 0.9.1
- **Arguments:** C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\main.py --user-directory C:\Users\shaun\Documents\ComfyUI\user --input-directory C:\Users\shaun\Documents\ComfyUI\input --output-directory C:\Users\shaun\Documents\ComfyUI\output --front-end-root C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\web_custom_versions\desktop_app --base-directory C:\Users\shaun\Documents\ComfyUI --database-url sqlite:///C:/Users/shaun/Documents/ComfyUI/user/comfyui.db --extra-model-paths-config C:\Users\shaun\AppData\Roaming\ComfyUI\extra_models_config.yaml --log-stdout --listen 127.0.0.1 --port 8000 --enable-manager
- **OS:** win32
- **Python Version:** 3.12.11 (main, Aug 18 2025, 19:17:54) [MSC v.1944 64 bit (AMD64)]
- **Embedded Python:** false
- **PyTorch Version:** 2.9.0+rocmsdk20251116
## Devices

- **Name:** cuda:0 AMD Radeon(TM) Graphics : native
- **Type:** cuda
- **VRAM Total:** 26956611584
- **VRAM Free:** 26798444544
- **Torch VRAM Total:** 0
- **Torch VRAM Free:** 0

## Logs

2026-01-26T11:03:59.625130 - Adding extra search path custom_nodes C:\Users\shaun\Documents\ComfyUI\custom_nodes
2026-01-26T11:03:59.625130 - Adding extra search path download_model_base C:\Users\shaun\Documents\ComfyUI\models
2026-01-26T11:03:59.625130 - Adding extra search path custom_nodes C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\custom_nodes
2026-01-26T11:03:59.625130 - Setting output directory to: C:\Users\shaun\Documents\ComfyUI\output
2026-01-26T11:03:59.625130 - Setting input directory to: C:\Users\shaun\Documents\ComfyUI\input
2026-01-26T11:03:59.625130 - Setting user directory to: C:\Users\shaun\Documents\ComfyUI\user
2026-01-26T11:03:59.705202 - [START] Security scan2026-01-26T11:03:59.705202 -
2026-01-26T11:04:00.357185 - [DONE] Security scan2026-01-26T11:04:00.357185 -
2026-01-26T11:04:00.358685 - ** ComfyUI startup time:2026-01-26T11:04:00.358685 - 2026-01-26T11:04:00.358685 - 2026-01-26 11:04:00.3582026-01-26T11:04:00.358685 -
2026-01-26T11:04:00.358685 - ** Platform:2026-01-26T11:04:00.358685 - 2026-01-26T11:04:00.358685 - Windows2026-01-26T11:04:00.358685 -
2026-01-26T11:04:00.358685 - ** Python version:2026-01-26T11:04:00.358685 - 2026-01-26T11:04:00.358685 - 3.12.11 (main, Aug 18 2025, 19:17:54) [MSC v.1944 64 bit (AMD64)]2026-01-26T11:04:00.358685 -
2026-01-26T11:04:00.358685 - ** Python executable:2026-01-26T11:04:00.358685 - 2026-01-26T11:04:00.358685 - C:\Users\shaun\Documents\ComfyUI\.venv\Scripts\python.exe2026-01-26T11:04:00.359184 -
2026-01-26T11:04:00.362196 - ** ComfyUI Path:2026-01-26T11:04:00.362196 - 2026-01-26T11:04:00.362196 - C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI2026-01-26T11:04:00.362196 -
2026-01-26T11:04:00.362196 - ** ComfyUI Base Folder Path:2026-01-26T11:04:00.362196 - 2026-01-26T11:04:00.362196 - C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI2026-01-26T11:04:00.362196 -
2026-01-26T11:04:00.362196 - ** User directory:2026-01-26T11:04:00.362196 - 2026-01-26T11:04:00.362196 - C:\Users\shaun\Documents\ComfyUI\user2026-01-26T11:04:00.362196 -
2026-01-26T11:04:00.362196 - ** ComfyUI-Manager config path:2026-01-26T11:04:00.362196 - 2026-01-26T11:04:00.362196 - C:\Users\shaun\Documents\ComfyUI\user\__manager\config.ini2026-01-26T11:04:00.362196 -
2026-01-26T11:04:00.362196 - ** Log path:2026-01-26T11:04:00.362196 - 2026-01-26T11:04:00.362196 - C:\Users\shaun\Documents\ComfyUI\user\comfyui.log2026-01-26T11:04:00.362196 -
2026-01-26T11:04:00.884102 - [ComfyUI-Manager] Skipped fixing the 'comfyui-frontend-package' dependency because the ComfyUI is outdated.
2026-01-26T11:04:00.884603 - [PRE] ComfyUI-Manager
2026-01-26T11:04:01.897763 - Checkpoint files will always be loaded safely.
2026-01-26T11:04:02.065342 - Total VRAM 25708 MB, total RAM 63093 MB
2026-01-26T11:04:02.065840 - pytorch version: 2.9.0+rocmsdk20251116
2026-01-26T11:04:02.065840 - Set: torch.backends.cudnn.enabled = False for better AMD performance.
2026-01-26T11:04:02.065840 - AMD arch: gfx1036
2026-01-26T11:04:02.065840 - ROCm version: (7, 1)
2026-01-26T11:04:02.066343 - Set vram state to: NORMAL_VRAM
2026-01-26T11:04:02.066343 - Device: cuda:0 AMD Radeon(TM) Graphics : native
2026-01-26T11:04:02.081860 - Using async weight offloading with 2 streams
2026-01-26T11:04:02.082357 - Enabled pinned memory 28391.0
2026-01-26T11:04:02.597296 - Found comfy_kitchen backend cuda: {'available': True, 'disabled': True, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8']}
2026-01-26T11:04:02.597296 - Found comfy_kitchen backend triton: {'available': False, 'disabled': True, 'unavailable_reason': "ImportError: No module named 'triton'", 'capabilities': []}
2026-01-26T11:04:02.597296 - Found comfy_kitchen backend eager: {'available': True, 'disabled': False, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8', 'scaled_mm_nvfp4']}
2026-01-26T11:04:02.813507 - Using sub quadratic optimization for attention, if you have memory or speed issues try using: --use-split-cross-attention
2026-01-26T11:04:04.041211 - Python version: 3.12.11 (main, Aug 18 2025, 19:17:54) [MSC v.1944 64 bit (AMD64)]
2026-01-26T11:04:04.041211 - ComfyUI version: 0.9.1
2026-01-26T11:04:04.071740 - [Prompt Server] web root: C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\web_custom_versions\desktop_app
2026-01-26T11:04:04.072240 - [START] ComfyUI-Manager
2026-01-26T11:04:04.174859 - [ComfyUI-Manager] network_mode: public
2026-01-26T11:04:04.176860 - [ComfyUI-Manager] The matrix sharing feature has been disabled because the `matrix-nio` dependency is not installed.
To use this feature, please run the following command:
C:\Users\shaun\Documents\ComfyUI\.venv\Scripts\python.exe -m pip install matrix-nio

2026-01-26T11:04:04.849540 -
❌ [Qwen3-TTS] Critical Import Error: No module named 'librosa'2026-01-26T11:04:04.849540 -
2026-01-26T11:04:04.849540 - Traceback for debugging:2026-01-26T11:04:04.849540 -
2026-01-26T11:04:04.851541 - Traceback (most recent call last):
2026-01-26T11:04:04.851541 - 2026-01-26T11:04:04.852040 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\nodes.py", line 79, in
import qwen_tts
2026-01-26T11:04:04.852040 - 2026-01-26T11:04:04.852040 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\qwen_tts\__init__.py", line 21, in
from .inference.qwen3_tts_model import Qwen3TTSModel, VoiceClonePromptItem
2026-01-26T11:04:04.852040 - 2026-01-26T11:04:04.852040 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\qwen_tts\inference\qwen3_tts_model.py", line 23, in
import librosa
2026-01-26T11:04:04.852540 - 2026-01-26T11:04:04.852540 - ModuleNotFoundError: No module named 'librosa'
2026-01-26T11:04:04.852540 - 2026-01-26T11:04:04.852540 -
During handling of the above exception, another exception occurred:

2026-01-26T11:04:04.852540 - 2026-01-26T11:04:04.853041 - Traceback (most recent call last):
2026-01-26T11:04:04.853041 - 2026-01-26T11:04:04.853041 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\nodes.py", line 85, in
from qwen_tts import Qwen3TTSModel, VoiceClonePromptItem
2026-01-26T11:04:04.853041 - 2026-01-26T11:04:04.853041 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\qwen_tts\__init__.py", line 21, in
from .inference.qwen3_tts_model import Qwen3TTSModel, VoiceClonePromptItem
2026-01-26T11:04:04.853540 - 2026-01-26T11:04:04.853540 - File "C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS\qwen_tts\inference\qwen3_tts_model.py", line 23, in
import librosa
2026-01-26T11:04:04.853540 - 2026-01-26T11:04:04.853540 - ModuleNotFoundError: No module named 'librosa'
2026-01-26T11:04:04.853540 - 2026-01-26T11:04:04.853540 -
Common fix: run 'pip install -r requirements.txt' in your ComfyUI environment.2026-01-26T11:04:04.853540 -
2026-01-26T11:04:04.854040 - ✅ ComfyUI-Qwen-TTS v1.0.1 loaded2026-01-26T11:04:04.854040 -
2026-01-26T11:04:04.855040 -
Import times for custom nodes:
2026-01-26T11:04:04.855542 - 0.0 seconds: C:\Users\shaun\AppData\Local\Programs\ComfyUI\resources\ComfyUI\custom_nodes\websocket_image_save.py
2026-01-26T11:04:04.855542 - 0.0 seconds: C:\Users\shaun\Documents\ComfyUI\custom_nodes\ComfyUI-Qwen-TTS
2026-01-26T11:04:04.855542 -
2026-01-26T11:04:04.859541 - Context impl SQLiteImpl.
2026-01-26T11:04:04.859541 - Will assume non-transactional DDL.
2026-01-26T11:04:04.872041 - Assets scan(roots=['models']) completed in 0.010s (created=0, skipped_existing=2, total_seen=2)
2026-01-26T11:04:04.894541 - Starting server

2026-01-26T11:04:04.895041 - To see the GUI go to: http://127.0.0.1:8000
2026-01-26T11:04:05.602495 - comfyui-frontend-package not found in requirements.txt
2026-01-26T11:04:07.076826 - [ComfyUI-Manager] The ComfyRegistry cache update is still in progress, so an outdated cache is being used.
2026-01-26T11:04:07.111328 - FETCH DATA from: C:\Users\shaun\Documents\ComfyUI\.venv\Lib\site-packages\comfyui_manager\custom-node-list.json2026-01-26T11:04:07.111828 - 2026-01-26T11:04:07.121327 - [DONE]2026-01-26T11:04:07.122329 -
2026-01-26T11:05:57.443033 - FETCH ComfyRegistry Data [DONE]
2026-01-26T11:05:57.560037 - [ComfyUI-Manager] default cache updated: https://api.comfy.org/nodes
2026-01-26T11:05:57.579539 - FETCH DATA from: C:\Users\shaun\Documents\ComfyUI\user\__manager\cache\1514988643_custom-node-list.json2026-01-26T11:05:57.579539 - 2026-01-26T11:05:57.587537 - [DONE]2026-01-26T11:05:57.588038 -
2026-01-26T11:05:57.615538 - [ComfyUI-Manager] All startup tasks have been completed.
2026-01-26T11:06:45.629099 - got prompt
2026-01-26T11:06:45.630099 - Failed to validate prompt for output 21:
2026-01-26T11:06:45.630099 - * LoadAudio 24:
2026-01-26T11:06:45.630099 - - Custom validation failed for node: audio - Invalid audio file: 雷军.mp3
2026-01-26T11:06:45.630099 - Output will be ignored
2026-01-26T11:06:45.633098 - WARNING: [Errno 2] No such file or directory: 'C:\\Users\\shaun\\Documents\\ComfyUI\\input\\雷军.mp3'
2026-01-26T11:06:45.635098 - ✅ [Qwen3-TTS] sdpa available (PyTorch built-in)2026-01-26T11:06:45.635098 -
2026-01-26T11:06:45.635098 - ✅ [Qwen3-TTS] eager attention available (always available)2026-01-26T11:06:45.635098 -
2026-01-26T11:06:45.635098 - 🔍 [Qwen3-TTS] Using requested attention: sdpa2026-01-26T11:06:45.635098 -
2026-01-26T11:06:45.635598 -
📥 [Qwen3-TTS] Downloading Qwen/Qwen3-TTS-12Hz-1.7B-VoiceDesign...2026-01-26T11:06:45.635598 -
2026-01-26T11:06:46.755905 -
Fetching 13 files: 0%| | 0/13 [00:00