报错：
ERROR DURING GENERATION: CUDA out of memory. Tried to allocate 72.00 MiB. GPU 0 has a total capacity of 12.00 GiB of which 6.88 GiB is free. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 52.20 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
==================================================
Traceback (most recent call last):
File "E:\AI\FramePack-V4\app.py", line 1141, in worker
File "E:\FramePack-V4\FramePack-V4\diffusers_helper\memory.py", line 92, in move_model_to_device_with_memory_preservation
if hasattr(m, 'weight'):
File "E:\FramePack-V4\FramePack-V4\diffusers_helper\memory.py", line 26, in hacked_get_attr
return torch.nn.Parameter(p.to(**kwargs), requires_grad=p.requires_grad)
torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 72.00 MiB. GPU 0 has a total capacity of 12.00 GiB of which 6.88 GiB is free. Of the allocated memory 2.69 GiB is allocated by PyTorch, and 52.20 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
yijing 发表于 2025-5-9 15:26
报错：
ERROR DURING GENERATION: CUDA out of memory. Tried to allocate 72.00 MiB. GPU 0 has a total...
环境变量PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True也设置了，没起作用
yijing 发表于 2025-5-9 15:26
报错：
ERROR DURING GENERATION: CUDA out of memory. Tried to allocate 72.00 MiB. GPU 0 has a total...
显存不足，推理内存保留设置大一些
无言以对 发表于 2025-5-9 16:05
显存不足，推理内存保留设置大一些
谢谢楼主，后来我尝试升级CUDA到12.6解决
https://img.picui.cn/free/2025/05/12/6821f533dff56.png 把下载解压好的文件用硬盘考到另一台电脑出现这个问题，
chenxi6553 发表于 2025-5-12 21:20
把下载解压好的文件用硬盘考到另一台电脑出现这个问题，
手动设置虚拟内存，32G起