search

antgroup / echomimic

EchoMimic: Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://antgroup.github.io/ai/echomimic/
Apache License 2.0
2.96k stars 343 forks source link

最低硬件要求多大? #139

Open yeohx opened 3 months ago

yeohx commented 3 months ago

我是阿里云服务器: 4核(vCPU) 30 GiB GPU：NVIDIA P100 启动 python -u webgui.py --server_port=3000 torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB. GPU 0 has a total capacity of 15.89 GiB of which 17.12 MiB is free. Process 1928 has 2.02 GiB memory in use. Process 2424 has 9.60 GiB memory in use. Including non-PyTorch memory, this process has 4.25 GiB memory in use. Of the allocated memory 3.86 GiB is allocated by PyTorch, and 142.06 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)

oisilener1982 commented 3 months ago

Echomimic is guaranteed working fine with RTX 3080 with 10gb. I can regularly generate 20+ seconds. It gave me a memory error when i tried to generate 3 minutes of audio source.

Try to limit the duration of your audio

nitinmukesh commented 2 months ago

Echomimic is guaranteed working fine with RTX 4060 with 8 GB VRAM + 8 GB shared RAM. I am regularly generating between 30 - 55 seconds without any issues.