| Index | index by Group | index by Distribution | index by Vendor | index by creation date | index by Name | Mirrors | Help | Search |
| Name: python314-comfy-aimdo | Distribution: openSUSE Tumbleweed |
| Version: 0.5.2 | Vendor: openSUSE |
| Release: 1.1 | Build date: Mon Sep 7 08:38:03 2026 |
| Group: Unspecified | Build host: reproducible |
| Size: 158836 | Source RPM: python-comfy-aimdo-0.5.2-1.1.src.rpm |
| Packager: https://bugs.opensuse.org | |
| Url: https://github.com/Comfy-Org/comfy-aimdo | |
| Summary: AI Model Dynamic Offloader for ComfyUI (pure-Python fallback) | |
AIMDO (AI Model Dynamic Offloader) is a PyTorch VRAM allocator used by ComfyUI to offload model weights under memory pressure. This build ships the pure-Python fallback. The native CUDA/HIP allocator is not compiled: it requires CUDA 12.8+ and a CUDA-enabled PyTorch, which Factory does not provide. Without the native module, init() returns False and ComfyUI continues without AIMDO.
GPL-3.0-only
* Mon Sep 07 2026 Martin Pluskal <mpluskal@suse.com>
- Update to 0.5.2:
* New malloc_graph_abort() to tear down an incomplete malloc graph
safely, aimed at a hard crash on workflow cancel; it lives in the
native allocator, which this build does not compile
* Wed Sep 02 2026 Martin Pluskal <mpluskal@suse.com>
- Update to 0.5.0:
* New "memory compiler": a native CUDA malloc-graph recorder that
replays an allocation trace instead of re-deciding it per step,
plus VBAR eviction during CUDA graph capture
* Adds the comfy_aimdo.malloc_graph module and control.record();
both need the native allocator, which this build does not compile
* Purely additive on the Python side -- no name was removed, and
every other module is byte-identical to 0.4.15, so consumers
pinning a >= floor are unaffected
* Ships a tests/ tree, but every case needs a CUDA device, so
%check stays the import/no-accelerator check
- Extend %check to import the new comfy_aimdo.malloc_graph module
* Tue Aug 25 2026 Martin Pluskal <mpluskal@suse.com>
- Update to 0.4.15:
* Windows shared-memory detection now keeps 512 MB of VRAM headroom
instead of the 768 MB introduced in 0.4.14 when the free-memory
figure is read from NVML rather than from cuMemGetInfo()
* That Windows-only constant is the entire release; the Python code
shipped by this package is unchanged from 0.4.14
* Sun Aug 23 2026 Martin Pluskal <mpluskal@suse.com>
- Stop warning "Could not autodetect AIMDO implementation, assuming
Nvidia" on every start: add
comfy-aimdo-detect-vendor-without-local-version.patch
* the vendor check looked for "+cu" or "+rocm" in the PyTorch
version string, which only upstream's own wheels carry, so a
distribution PyTorch never matched and the code then guessed
Nvidia; it now reads the cuda and hip attributes that the same
version module already defines
* when PyTorch has neither, there is no allocator to load, so
init() reports failure with an informational message instead of
warning about a vendor it never really selected
* %check now exercises that path
* Sun Aug 23 2026 Martin Pluskal <mpluskal@suse.com>
- Update to 0.4.14:
* README now documents AMD GPU (ROCm 7+) support alongside Nvidia
(CUDA 12.8+); the HIP backend itself already shipped in 0.4.13
and is unchanged in this release
* Windows shared-memory detection reserves a larger VRAM headroom
(768 MB) when the free-memory figure comes from NVML rather than
from cuMemGetInfo()
* No changes to the pure-Python fallback shipped by this package
* Fri Aug 21 2026 Martin Pluskal <mpluskal@suse.com>
- Initial package of python-comfy-aimdo 0.4.13:
* Pure-Python fallback; native CUDA/HIP allocator is not built
(needs CUDA 12.8+ and a CUDA-enabled PyTorch)
* Require python-torch >= 2.8.0
/usr/lib/python3.14/site-packages/comfy_aimdo /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/INSTALLER /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/METADATA /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/RECORD /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/REQUESTED /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/WHEEL /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/licenses /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/licenses/LICENSE /usr/lib/python3.14/site-packages/comfy_aimdo-0.5.2.dist-info/top_level.txt /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__ /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/_version.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/_version.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/control.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/control.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/host_buffer.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/host_buffer.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/malloc_graph.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/malloc_graph.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/model_mmap.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/model_mmap.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/model_vbar.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/model_vbar.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/torch.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/torch.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/vram_buffer.cpython-314.opt-1.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/__pycache__/vram_buffer.cpython-314.pyc /usr/lib/python3.14/site-packages/comfy_aimdo/_version.py /usr/lib/python3.14/site-packages/comfy_aimdo/control.py /usr/lib/python3.14/site-packages/comfy_aimdo/host_buffer.py /usr/lib/python3.14/site-packages/comfy_aimdo/malloc_graph.py /usr/lib/python3.14/site-packages/comfy_aimdo/model_mmap.py /usr/lib/python3.14/site-packages/comfy_aimdo/model_vbar.py /usr/lib/python3.14/site-packages/comfy_aimdo/torch.py /usr/lib/python3.14/site-packages/comfy_aimdo/vram_buffer.py /usr/share/doc/packages/python314-comfy-aimdo /usr/share/doc/packages/python314-comfy-aimdo/README.md /usr/share/licenses/python314-comfy-aimdo /usr/share/licenses/python314-comfy-aimdo/LICENSE
Generated by rpm2html 1.8.1
Fabrice Bellet, Wed Sep 9 22:22:31 2026