修改缓存与执行顺序以最小化RAM占用：节点输出在不再需要时立即从缓存删除，尽量减少同时需要缓存的输出，从而在内存受限时保证工作流可运行，但会显著降低执行速度并导致重复执行。
💡 在内存受限的环境下运行大型ComfyUI工作流以避免内存不足
🍴 2 Forks💻 Python🔄 2024-12-25
https://pan.quark.cn/s/e00a65475347
📄 README
ComfyUI-mincache
Modifies the caching and execution order to minimize RAM usage
Node outputs are deleted from the cache as soon as they are no longer needed. Thus, every node re-executes on every queue.
Execution is modified to try and minimize the number of outputs which need to be cached at once.
At current, expect to have severely worse execution time unless the workflow being used would otherwise fail to execute.