Skip to content

Pull requests: NVIDIA/Model-Optimizer

New pull request New
236 Open 1,346 Closed
236 Open 1,346 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Add sidecar GPU/CPU memory+utilization monitor for HF PTQ
#2000 opened Jul 21, 2026 by Fridah-nv Contributor • Draft
3
Speed up compressed-tensors load-time matching (for Kimi models)
#1999 opened Jul 21, 2026 by rohansjoshi Contributor Loading…
6
Offline-KD QAD example
#1998 opened Jul 20, 2026 by AAnoosheh Contributor Loading…
2
fix(export): prevent silent MLP linear collapse during HF export
#1997 opened Jul 20, 2026 by Surya-5555 Loading…
3
Add link to puzzletron_v2
#1996 opened Jul 20, 2026 by Separius Contributor Loading…
7
fix(autoquant): score grouped QKV at attention output
#1993 opened Jul 19, 2026 by realAsma Contributor • Draft
8
Adds skip-softmax calibration through the vLLM serving path
#1992 opened Jul 19, 2026 by kaix-nv Contributor • Draft
4
Document non-Diffusers FP8/NVFP4 ComfyUI export
#1991 opened Jul 17, 2026 by jingyu-ml Contributor • Draft
3
feat(rocm): Add AMD ROCm/MI300X support — FP8 hipBLASLt, MIGraphX backend, AMD quantization configs
#1990 opened Jul 17, 2026 by zhihuidu-amd Loading…
5
[6008361][ONNX][Quantization] Clarify autotune guidance
#1989 opened Jul 17, 2026 by gcunhase Contributor • Draft
3
Make group-boundary AutoQuant scoring the default
#1988 opened Jul 17, 2026 by meenchen Contributor • Draft
4
Add MLflow tracking to modelopt MCP
#1986 opened Jul 17, 2026 by ChenhanYu Collaborator Loading…
@ChenhanYu
18
[6425069][ONNX][Autocast] Fix autocast metadata propagation
#1983 opened Jul 16, 2026 by gcunhase Contributor Loading…
13
launcher: bump TRT-LLM to 1.3.0rc20, pin vLLM to v0.22.0, fix max_seq…
#1982 opened Jul 16, 2026 by noeyy-mino Contributor Loading…
5
[5726458] Add NVFP4 projection-output-quantizer recipe and HF embedding ONNX export example
#1981 opened Jul 16, 2026 by ajrasane Contributor Loading…
20
Scripts and a skill to do per-layer benchmark using flashinfer
#1980 opened Jul 16, 2026 by sychen52 Contributor Loading…
40
docs: add AutoQuantize mixed-precision search blog
#1979 opened Jul 15, 2026 by realAsma Contributor • Draft
30
feat(specdec): warm-start DFlash fine-tuning from exported drafter via dflash_init_checkpoint
#1978 opened Jul 15, 2026 by h-guo18 Contributor • Draft
4
Quality: Insecure subprocess usage in get_system_info.py
#1977 opened Jul 15, 2026 by tomaioo Loading…
9
Add activation-aware local Hessian scale calibration
#1976 opened Jul 14, 2026 by realAsma Contributor • Draft
20
add Qwen3-VL support for DFlash training
#1975 opened Jul 14, 2026 by skierat Contributor Loading…
12
Add MoE support to runtime stats
#1973 opened Jul 14, 2026 by grzegorz-k-karch Contributor • Draft
4
Multi-GPU vLLM benchmarking for runtime stats
#1972 opened Jul 14, 2026 by grzegorz-k-karch Contributor • Draft
4
docs: add announcements landing page
#1971 opened Jul 13, 2026 by ChenhanYu Collaborator • Draft
4
[Examples]: MiniMax-M3 DSpark
#1965 opened Jul 12, 2026 by h-guo18 Contributor Loading…
11
ProTip! no:milestone will show everything without a milestone.