-
![火狐截图_2024-06-18T12-21-01 778Z](https://github.com/AIFSH/ComfyUI-MuseTalk_FSH/assets/171790958/c8432492-aa9b-4519-a700-119100532009)
您好:下面的文件中有错误,请查看:
=====comfyui-musetalk_fsh/inference.py的36行…
-
我用的是musetalk,当我打开webrtcchat网页,第一次点击start按钮与数字人对话一切正常,点击stop后刷新网页,再次点击start,fps就会降低,asr变卡顿;重复上述操作,fps越来越低。
-
tesla m40 24G,text2video,生成3秒需要1个多小时,但是将产生的这个视频用MuseTalk处理很快。MuseV是没优化好,还是过期显卡就这个效果?还是设置问题?
-
# ComfyUI Error Report
## Error Details
- **Node Type:** UNETLoader_MuseTalk
- **Exception Type:** AttributeError
- **Exception Message:** 'MuseModelConfig' object has no attribute 'custom_operati…
-
i have rtx 3060ti with 8gb vram
[rgthree] Using rgthree's optimized recursive execution.
model_type EPS
!!! Exception during processing!!! 'MuseModelConfig' object has no attribute 'memory_usage_…
-
Most of us heard about the musetalk is photo talking...
Why the sample you provided is a video ???
-
您好,请问musetalk是无法实现rtmp推送还是还没有做这一块呢
-
您好,我写了一个onnx导出脚本,只导出unet.model,然而导出后文件并不是保存在一个model.onnx中,,而是model.onnx只保存文件结构,而权重保存成零散的文件?
导出代码如下:
```
# ===============================构建算子
import onnxscript
## Assuming you use opset18
fr…
-
/root/anaconda3/envs/muse/bin/python 7624MiB
启动运行过后 gpu一直在7g左右,请问哪里可以优化释放内存. 里面unet 和 vae 用的么?
-
我在32G显存的v100下跑实时推理示例:python -m scripts.realtime_inference --inference_config configs/inference/realtime.yaml --batch_size 4
得到的结果是“Total process time of 202 frames including saving images = 25.9267…