Search results

NVFP4 __Dyna layer took most of the inference time

bugSomething isn't workingSomething isn't working

Status: Open.

#751 In NVIDIA/Model-Optimizer;

· CHOcho-quan opened on Jan 8, 2026

NVFP4 Quantization for GLM 4.7 w MTP

bugSomething isn't workingSomething isn't working

Status: Open.

#750 In NVIDIA/Model-Optimizer;

· pdasgup opened on Jan 8, 2026

Need guidance on converting ModelOpt QAT model (ResNet50) to ONNX then TensorRT

questionHelp is is neededHelp is is needed

Status: Open.

#741 In NVIDIA/Model-Optimizer;

· Sam-gsj opened on Jan 7, 2026

When try to prune a Qwen2.5 vl model, I got error: Only rope position embedding is supported, got mrope

feature requestNew feature or requestNew feature or request

torch.pruning

Status: Open.

#739 In NVIDIA/Model-Optimizer;

· jackmego opened on Jan 6, 2026

MTP quantizing doesn't work - Qwen3 Next

bugSomething isn't workingSomething isn't working

model support

torch.quantization

waiting for feedback

Status: Open.

#733 In NVIDIA/Model-Optimizer;

· nutriarch opened on Dec 28, 2025

[Doc - Guidance] Proper MoE quantization

documentationImprovements or additions to documentationImprovements or additions to documentation

feature requestNew feature or requestNew feature or request

torch.quantization

waiting for feedback

Status: Open.

#732 In NVIDIA/Model-Optimizer;

· mratsim opened on Dec 28, 2025

Issue: Weight-Only Quantization Not Working for GroupedLinear Modules

featureIssue in actual ModelOpt featureIssue in actual ModelOpt feature

questionHelp is is neededHelp is is needed

torch.quantization

waiting for feedback

Status: Open.

#725 In NVIDIA/Model-Optimizer;

· jQizhang opened on Dec 23, 2025

Issues with torch_quant_to_onnx.py

bugSomething isn't workingSomething isn't working

installationIssue related to Installing ModelOptIssue related to Installing ModelOpt

onnx.quantization

waiting for feedback

Status: Open.

#724 In NVIDIA/Model-Optimizer;

· HaohuiHU opened on Dec 23, 2025

Quantizing Qwen VL 30bA3

model support

questionHelp is is neededHelp is is needed

torch.quantization

waiting for feedback

Status: Open.

#716 In NVIDIA/Model-Optimizer;

· bhaktatejas922 opened on Dec 21, 2025

Fail to convert ONNX models with dynamic inputs

bugSomething isn't workingSomething isn't working

featureIssue in actual ModelOpt featureIssue in actual ModelOpt feature

onnx.quantization

waiting for feedback

Status: Open.

#713 In NVIDIA/Model-Optimizer;

· jricker2 opened on Dec 19, 2025

How to add the original student model loss to the model distillation demo

questionHelp is is neededHelp is is needed

torch.distillation

waiting for feedback

Status: Open.

#712 In NVIDIA/Model-Optimizer;

· MeteorQ opened on Dec 19, 2025

CUDA Out of memory when mqt.compress model after QAT

bugSomething isn't workingSomething isn't working

waiting for feedback

Status: Open.

#696 In NVIDIA/Model-Optimizer;

· alrolo3 opened on Dec 15, 2025