Skip to content

Pull requests: NVIDIA/Model-Optimizer

New pull request New
316 Open 1,744 Closed
316 Open 1,744 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

fix(speculative): make streaming hidden-state capture work on vLLM 0.29
#2522 opened Sep 23, 2026 by h-guo18 Contributor Loading…
6
[2/4] GDN/KDA decode QAT with INT8 recurrent state
#2519 opened Sep 23, 2026 by kaix-nv Contributor • 2/4 • Draft
3
Fix vLLM compilation guard for models without marker
#2518 opened Sep 23, 2026 by kinjalpatel27 Contributor Loading…
7
[OMNIML-5899] Export Q8_0 checkpoints and add recipes
#2517 opened Sep 22, 2026 by hychiang-git Contributor Loading…
4
[OMNIML-5899] Add Q8_0 quantization codec and backend
#2516 opened Sep 22, 2026 by hychiang-git Contributor Loading…
5
[OMNIML-5899] Add Q8_0 CUDA packing kernel
#2515 opened Sep 22, 2026 by hychiang-git Contributor Loading…
5
Track the distillation and export stages of a Megatron-Bridge run
#2514 opened Sep 22, 2026 by kevalmorabia97 Collaborator Loading…
45
Add the IQ1_M weight-only quantization format
#2513 opened Sep 22, 2026 by cjluo-nv Collaborator Loading…
5
Add the IQ2_S weight-only quantization format
#2512 opened Sep 22, 2026 by cjluo-nv Collaborator Loading…
3
Add the IQ2_XXS weight-only quantization format
#2511 opened Sep 22, 2026 by cjluo-nv Collaborator Loading…
22
Use HF converter.rename_source_key when available to reverse naming conversion
#2508 opened Sep 22, 2026 by meenchen Contributor Loading…
15
[4/4] Experimental GDN/KDA approximate-inverse QAT
#2507 opened Sep 22, 2026 by kaix-nv Contributor • 4/4 • Draft
3
[3/4] GDN/KDA prefill operand and arithmetic QAT
#2503 opened Sep 22, 2026 by kaix-nv Contributor • 3/4 • Draft
3
Fix grouped expert quantizer checkpoint replicas
#2500 opened Sep 22, 2026 by hychiang-git Contributor Loading…
5 tasks done
8
feat(distill): support padded vocabulary tails
#2498 opened Sep 22, 2026 by efegokdemir Loading…
1
4
[1/4] GDN state/W QAT foundation
#2497 opened Sep 22, 2026 by kaix-nv Contributor • 1/4 • Draft
4
specdec_bench: forward kv_cache_dtype to the vLLM engine
#2496 opened Sep 22, 2026 by yeyu-nvidia Contributor Loading…
7
Make kernels.common.attention independent of kernels.sparsity
#2495 opened Sep 22, 2026 by sychen52 Contributor • Draft
3
Fix GitLab authentication guidance for Enroot
#2493 opened Sep 21, 2026 by chadvoegele Contributor • Draft
3
Bound the stale-capture warning to once per capture and name its cause
#2492 opened Sep 21, 2026 by Edwardssss Loading…
1
7
Use math.fsum for StaticLossBalancer weight checks
#2491 opened Sep 21, 2026 by aniruddh-krovvidi Loading…
1
2
fix: use explicit head dim for Puzzletron KV cache
#2485 opened Sep 20, 2026 by devxMani Loading…
1
10
[Puzzletron] Fix KV-cache dimension calculation for models with custom head_dim (#2482)fix(puzzletron): respect explicit head_dim in KV-cache dimension calc…
#2484 opened Sep 20, 2026 by musi22 Loading…
2
docs(eval): align Terminal-Bench 2.1 / SWE-bench / MRCR with upstream configs
#2479 opened Sep 18, 2026 by cjluo-nv Collaborator Loading…
4
Add MLflow tracking flags to megatron_bridge quantize.py
#2477 opened Sep 18, 2026 by kevalmorabia97 Collaborator Loading…
52
ProTip! Exclude everything labeled bug with -label:bug.