RuntimeError: [enforce fail at alloc_cpu.cpp:115] data. DefaultCPUAllocator: not enough memory: you tried to allocate X bytes.

The specific error is:
Traceback (most recent call last):
File "S:\framepack_cu126_torch26\webui\demo_gradio.py", line 298, in worker
save_bcthw_as_mp4(history_pixels, output_filename, fps=30, crf=mp4_crf)
File "S:\framepack_cu126_torch26\webui\diffusers_helper\utils.py", line 276, in save_bcthw_as_mp4
x = torch.clamp(x.float(), -1., 1.) * 127.5 + 127.5
RuntimeError: [enforce fail at alloc_cpu.cpp:115] data. DefaultCPUAllocator: not enough memory: you tried to allocate 7596711936 bytes.

at least the last time it failed. Note that it did fail in exactly the same way using the same image with the same 54 second mark so might be specific to this particular image

It seems to happen when I hit 54 seconds of video constantly. I have 48GB RAM (yeah I'm one of those) and about 84GB total swap space set just in case. Watching the RAM usage and Swap it doesn't max out. In fact only about 34Gb RAM is in use when this error shows (unless it spikes right before an I miss it)

Not sure what all you might need from this but I'm using Windows 11 etc.

Reproduce the failure around 54 seconds using the reported image, then inspect webui/demo_gradio.py at line 298 and webui/diffusers_helper/utils.py at line 276. Track the size of history_pixels and the allocation made during save_bcthw_as_mp4; done means identifying the cause of the memory failure and documenting or validating a fix.

Written by the indexing model from the issue text.