Skip to content

Pull requests: NVIDIA/Model-Optimizer

New pull request New
120 Open 663 Closed
120 Open 663 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 DeepSeek MoE detection and export mapping in HF PTQ/export path
#1125 opened Mar 26, 2026 by Charles-JCJ Loading…
3
Refine _extract_layer_prefixes to better handle mtp modules
#1124 opened Mar 26, 2026 by Edwardf0t1 Loading…
5
Remove --skip-softmax option from hf_sa (only allow calibration option)
#1123 opened Mar 25, 2026 by rohansjoshi Loading…
3
Remove custom DistillationProvider and simplify mbridge distillation and hf export
#1122 opened Mar 25, 2026 by kevalmorabia97 Loading…
7
Merge puzzletron compression algorithm
#1121 opened Mar 25, 2026 by danielkorzekwa Loading…
16
Added fallback to preload cudnn dlls from nvidia cudnn venv package or torch venv package
#1119 opened Mar 25, 2026 by hthadicherla Loading…
7
Add custom MoE quantization guide for HuggingFace models
#1118 opened Mar 25, 2026 by cjluo-nv Loading…
3
[BugFix][5271237][ONNX] Add Q/DQ placement for Conv->LayerNorm patterns
#1117 opened Mar 24, 2026 by ajrasane Loading…
@ajrasane
9
[OMNIML-1525] Create a folder for the plugin example
#1114 opened Mar 24, 2026 by ajrasane Loading…
29
Add nvfp4_mse and nvfp4_local_hessian options to the ptq script. cherry-pick After code freeze, cherry-pick into release branch for next rc. Only for bug fixes and doc updates
#1113 opened Mar 24, 2026 by bkartal-dev Loading…
6
Add bypass distillation (blockwise local KD) to puzzletron pipeline
#1111 opened Mar 24, 2026 by Separius Loading…
11
Add Agent PTQ skill for model quantization
#1107 opened Mar 24, 2026 by mxinO • Draft
19
[OMNIML-3776]: add clear docs restrict the model types
#1105 opened Mar 23, 2026 by shengliangxu Loading…
4
fix: EAGLE mix_hidden_states in-place op crash (#1088)
#1104 opened Mar 23, 2026 by javierdejesusda Loading…
6 tasks done
1
3
Run hf ptq on computelab
#1100 opened Mar 23, 2026 by cjluo-nv • Draft
5
Added general graph surgery run function for easier scalability with Olive.
#1096 opened Mar 23, 2026 by hthadicherla Loading…
8
[OMNIML-3689] PTQ quant_cfg semantic correction. Design in doc _quant_cfg.rst
#1094 opened Mar 22, 2026 by shengliangxu Loading…
35
Exclude small-channel Conv nodes from FP8 quantization
#1083 opened Mar 20, 2026 by nv-samcheng Loading…
16
Unified diffusion model trainer
#1082 opened Mar 20, 2026 by mxinO • Draft
3
[3/n] Add skip-softmax to Triton flash attention kernel
#1081 opened Mar 20, 2026 by kaix-nv Loading…
30
fix: [modelopt 0.43.0][GB200][llm_ptq / sglang] Llama-3.1-8B-Inst (#5997673)
#1080 opened Mar 20, 2026 by ChenhanYu Loading…
3
fix: [modelopt 0.43][GH200][llm_ptq - autoquant / trtllm] Llama-3 (#5997832)
#1079 opened Mar 20, 2026 by ChenhanYu • Draft
4
fix: [ModelOpt-Windows][modelopt 0.43.0] [genai_llm][README]: Sho (#5997787)
#1077 opened Mar 19, 2026 by ChenhanYu Loading…
6
fix: Speculative Decoding (#1066)
#1076 opened Mar 19, 2026 by ChenhanYu • Draft
1
6
fix: Feature: Add validation for loaded modelopt state files (#1041)
#1074 opened Mar 19, 2026 by ChenhanYu Loading…
1
10
ProTip! no:milestone will show everything without a milestone.