trending/Model-Optimizer
1
0
Fork 0
mirror of https://github.com/NVIDIA/Model-Optimizer.git synced 2026-11-13 02:17:13 +08:00
Code Issues Packages Projects Releases Wiki Activity
121 Commits 503 Branches 62 Tags
00a7e6079f87d1ca9862da605bf47b0245c04ae4
Commit Graph
57 Commits
Author SHA1 Message Date
+2
yeyu-nvidiakinjalpatel27Asha AnooshehrealAsmaChenjie LuoRiyad Islamyueshen2016h-guo18Chenjie LuoKeval MorabiaomrialmogJenny Chen
 		00a7e6079f remove DetachedEagleGPT model and handle all offline mode in the _DynamicEagleGPTModel (#321) 
Signed-off-by: Ye Yu <yeyu@nvidia.com>
Signed-off-by: Kinjal Patel <kinjalpravin@nvidia.com>
Signed-off-by: Asha Anoosheh <aanoosheh@nvidia.com>
Signed-off-by: realAsma <akuriparambi@nvidia.com>
Signed-off-by: Chenjie Luo <chenjiel@nvidia.com>
Signed-off-by: Riyad Islam <rislam@nvidia.com>
Signed-off-by: Yue <yueshen@nvidia.com>
Signed-off-by: h-guo18 <67671475+h-guo18@users.noreply.github.com>
Signed-off-by: Chenjie Luo <108829653+cjluo-nv@users.noreply.github.com>
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
Signed-off-by: omrialmog <oalmog@nvidia.com>
Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>
Co-authored-by: kinjalpatel27 <31936134+kinjalpatel27@users.noreply.github.com>
Co-authored-by: Asha Anoosheh <aanoosheh@nvidia.com>
Co-authored-by: realAsma <86726418+realAsma@users.noreply.github.com>
Co-authored-by: Chenjie Luo <108829653+cjluo-nv@users.noreply.github.com>
Co-authored-by: Riyad Islam <rislam@nvidia.com>
Co-authored-by: yueshen2016 <39203804+yueshen2016@users.noreply.github.com>
Co-authored-by: h-guo18 <67671475+h-guo18@users.noreply.github.com>
Co-authored-by: Chenjie Luo <chenjie@omniml.ai>
Co-authored-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
Co-authored-by: omrialmog <oalmog@nvidia.com>
Co-authored-by: Jenny Chen <jennifchen@nvidia.com>
 2025-09-18 22:48:03 +00:00
Asha Anoosheh b895dc5c7c Disable KD mode from saving problematic state (#320) 
Signed-off-by: Asha Anoosheh <aanoosheh@nvidia.com>
 2025-09-19 02:31:20 +05:30
ajrasane 27a68219ee Specify engine path for TRT evaluation (#313) 
Signed-off-by: ajrasane <131806219+ajrasane@users.noreply.github.com>
 2025-09-18 13:09:53 -07:00
Jenny Chen 1c7c16e0c8 Slurm support for QAT Simplified Flow + Qwen3-8B recipe (#285) 
Signed-off-by: Jennifer Chen <jennifchen@nvidia.com>
 2025-09-18 11:24:39 -07:00
yueshen2016 bbb2304491 Fix issue of attention.core_attention is None (#334) 
Signed-off-by: Yue <yueshen@nvidia.com>
 2025-09-18 01:19:51 +00:00
Chenjie LuoChenjie LuoKeval Morabia
 		146d1d9d40 Deprecate TRTLLM-build in examples (#297) 
Signed-off-by: Chenjie Luo <chenjiel@nvidia.com>
Signed-off-by: Chenjie Luo <108829653+cjluo-nv@users.noreply.github.com>
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
Co-authored-by: Chenjie Luo <chenjie@omniml.ai>
Co-authored-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-18 03:22:37 +05:30
Asha Anoosheh 5db7169887 Allow KD loss in val mode for MLM plugin (#331) 
Signed-off-by: Asha Anoosheh <aanoosheh@nvidia.com>
 2025-09-17 21:52:02 +00:00
yueshen2016 d94fc1b44e Fix issues of attention.core_attention.softmax_offset is None (#330) 
Signed-off-by: Yue <yueshen@nvidia.com>
 2025-09-17 20:25:38 +00:00
Riyad Islam 682bf6d37d import fix for torch 2.9 (#315) 
Signed-off-by: Riyad Islam <rislam@nvidia.com>
 2025-09-17 04:21:06 +00:00
realAsma 35247328d7 [1/N] QATTrainer training workflow fixes and clean up; Added backend specific unitests; (#318) 
Signed-off-by: realAsma <akuriparambi@nvidia.com>
 2025-09-17 01:43:44 +00:00
realAsma 8a5736aea9 Remove unused utilities for ModelOpt <0.29 MCore checkpoints (#322) 
Signed-off-by: realAsma <akuriparambi@nvidia.com>
 2025-09-16 12:45:44 -07:00
Asha Anoosheh 9aedfdf71a Update distill Megatron plugin (#319) 
Signed-off-by: Asha Anoosheh <aanoosheh@nvidia.com>
 2025-09-16 12:56:23 +02:00
kinjalpatel27 d406aa10e2 Added support for qwen3-next quantization and export (#323) 
Signed-off-by: Kinjal Patel <kinjalpravin@nvidia.com>
 2025-09-16 12:06:50 +05:30
Gwena Cunha 4ea72e3e20 [5504719] Fix bypassing of 'Cast' connecting a consumer with multiple outputs a… (#309) 
Signed-off-by: gcunhase <4861122+gcunhase@users.noreply.github.com>
 2025-09-10 21:43:47 +00:00
Riyad Islam 4716131196 Avoid autocast at onnx export if fp32 model is desired (#304) 
Signed-off-by: Riyad Islam <rislam@nvidia.com>
 2025-09-10 16:28:37 +00:00
Riyad Islam 85b309f15e Qwen quantize and hf export support in examples (#311) 
Signed-off-by: Riyad Islam <rislam@nvidia.com>
 2025-09-10 07:11:10 +00:00
Riyad Islam b913290dfd Making stronglyTyped default for modelopt evaluation (#287) 
Signed-off-by: Riyad Islam <rislam@nvidia.com>
 2025-09-10 02:19:26 +00:00
ajrasane 0adb1b6772 Upgrade to ONNX 1.19.0 (#289) 
Signed-off-by: ajrasane <131806219+ajrasane@users.noreply.github.com>
 2025-09-10 00:40:39 +00:00
Gwena Cunha d6d2e75b63 [5452146] Fix: 'Invalid tensor data type 0' (#308) 
Signed-off-by: gcunhase <4861122+gcunhase@users.noreply.github.com>
 2025-09-09 14:04:53 -04:00
h-guo18 a6fa34cda4 Feat: update eagle3 example; add export (#293) 
Signed-off-by: h-guo18 <67671475+h-guo18@users.noreply.github.com>
 2025-09-08 18:02:41 -07:00
Chenjie Luo 512dbb7a08 Avoid squeezing original weight tensors with leading dim 1 (#294) 
Signed-off-by: Chenjie Luo <108829653+cjluo-nv@users.noreply.github.com>
 2025-09-09 00:12:11 +00:00
Gwena Cunha d5c88e7a27 [5477976] Fix: issue removing Q/DQ nodes around custom ops with constant inputs (#296) 
Signed-off-by: gcunhase <4861122+gcunhase@users.noreply.github.com>
 2025-09-05 23:56:39 +00:00
Chenhan D. Yu 67af656f5b Fixing multi-layer eagle3 forward hook placement (#295) 
Signed-off-by: Chenhan Yu <chenhany@nvidia.com>
 2025-09-05 15:53:21 -07:00
Keval Morabia 1cf78b2231 Remove pruning supported model list in code (#290) 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-05 22:57:30 +05:30
yeyu-nvidia 8a07376863 set pipeline_dtype default value to params_dtype in megatron_eagle TransformerConfig (#291) 
Signed-off-by: Ye Yu <yeyu@nvidia.com>
 2025-09-05 10:07:25 -07:00
Keval Morabia 76fb12d476 Move pruning nemo import in function to avoid circular import 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-04 22:14:26 -07:00
Zhiyu 78008ac59c Make vlm detection more robust in ptq workflow (#286) 
Signed-off-by: Zhiyu Cheng <zhiyuc@nvidia.com>
 2025-09-04 19:53:17 -07:00
Riyad IslamandKeval Morabia d0372f4401 Training mode removal from ONNX nodes (#277) 
Signed-off-by: Riyad Islam <rislam@nvidia.com>
Co-authored-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-04 11:27:57 -07:00
Keval Morabia 103b1bb137 Push latest changes 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-04 11:02:58 +05:30
Keval Morabia 1ef1d72a1b Code quality improvements - typos, formatting, etc. 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-09-02 19:58:33 +05:30
Keval Morabia c391942107 Move triton import inside function to work on Windows 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-30 11:40:09 +05:30
Keval Morabia 4d1eb0caf5 Major improvement of READMEs and documentation 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-30 10:54:28 +05:30
Keval Morabia d09c82779c Push latest changes and bug fixes (#274) 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-29 17:29:52 +05:30
Keval Morabia dcdc2df084 Push latest changes 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-26 01:12:26 +05:30
Keval Morabia 37085354c9 Update files on Github (#258) 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-17 02:44:16 +05:30
Keval Morabia 0bea1c31d7 Add extensive CI unit tests + torch 2.8 fix (#244) 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-08-08 11:56:47 +05:30
Keval Morabia 4c611e47a6 Update files on GitHub 
Signed-off-by: Keval Morabia <28916987+kevalmorabia97@users.noreply.github.com>
 2025-07-31 22:48:09 +05:30
Keval Morabiaandnamgyu-youn 7a27f2a59b Update Ln-norm logic for upcoming PyTorch update (#206) 
Co-authored-by: namgyu-youn <yynk2012@gmail.com>
 2025-07-14 23:19:04 +05:30
Keval Morabiaandishan-modi 8e3bfb5703 Fix NF4 scale padding (#183) 
Co-authored-by: ishan-modi <ishan.modi24@gmail.com>
 2025-07-14 23:17:58 +05:30
Keval Morabia cafa7f60ed Update for 0.33.0 release 2025-07-14 23:08:26 +05:30
Keval MorabiaandMichael Feil cdc8996820 Qwen3 TRT-LLM/HF export (#188) 
Co-authored-by: Michael Feil <63565275+michaelfeil@users.noreply.github.com>
 2025-06-05 13:25:28 -07:00
Keval Morabia 7af33d29ce Update for 0.31.0 release 2025-06-05 13:24:07 -07:00
Keval Morabia 9c54aa1c47 Fix license in export_onnx.py (!195) 2025-05-16 22:23:16 +05:30
Keval Morabia acecdb5ba6 Fix svdquant restore 2025-05-09 09:37:59 +05:30
Keval Morabiaandishan-modi bf2b0d028b Add INT8 realquant support (#166) 
Co-authored-by: ishan-modi <ishan.modi24@gmail.com>
 2025-05-08 23:46:15 +05:30
Keval Morabia 7a047435ae Update for 0.29.0 release 2025-05-08 23:43:51 +05:30
Keval Morabia e048fb239d Add changes for 0.27 Windows release 2025-04-29 22:11:52 +05:30
Keval Morabiaandmichaelfeil d59ca04470 Support Qwen MOE with the TRT LLM C++ route (#155) 
Co-authored-by: michaelfeil <63565275+michaelfeil@users.noreply.github.com>
 2025-04-15 23:39:37 +05:30
Keval Morabia 9cb36cb43c Update files for 0.27.1 release 2025-04-15 23:36:35 +05:30
Keval Morabiaandmichaelfeil 7b462dd278 Added check of the output names while removing cast nodes (#137) 
Co-Authored-By: michaelfeil <63565275+michaelfeil@users.noreply.github.com>
 2025-04-03 10:45:37 +05:30