Skip to content

Issues: NVIDIA/TensorRT-Model-Optimizer

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

61 Open 51 Closed
61 Open 51 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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Quantization] activation bitwidth for memory access
#119 opened Dec 22, 2024 by yokosyun
quantize resnet18 onnx with fp8
#118 opened Dec 20, 2024 by thishome
[ONNX][PTQ] Quantization failed with --dq_only flag in ConvTranspose
#117 opened Dec 19, 2024 by ry3s
Asymmetric quantization for Activation?
#115 opened Dec 15, 2024 by yokosyun
What is difference of torch.quantization and onnx.quantization for speed and accuracy ?
#114 opened Dec 11, 2024 by demuxin
Is there a plan to support more recent PTQ methods for INT8 ViT?
#113 opened Dec 10, 2024 by dedoogong
Question aboutf use QuantizeLinear Node with my custom op
#112 opened Dec 10, 2024 by AnnaTrainingG
AssertionError: We only support fp8 for SDXL on Level 4
#109 opened Nov 25, 2024 by wxsms
@jingyu-ml
[RFC] TensorRT Model Optimizer - Product Roadmap roadmap
#108 opened Nov 21, 2024 by hchings
6
how to reduce memory usage?
#107 opened Nov 20, 2024 by dedoogong
@riyadshairi979
1
FP16 and FP32 shows 30% lower accuracy compared to INT8 for the ViT Example in ONNX_PTQ
#106 opened Nov 13, 2024 by chjej202
@riyadshairi979
All the modules are disabled and onnx exporation is failed
#104 opened Nov 10, 2024 by Worromots
@jingyu-ml
Whether fp8 quantization supports the the DIT module？
#102 opened Nov 4, 2024 by Rudin6
1
In cache_diffusion example, can we use dynamic image shape & batch size?
#101 opened Nov 4, 2024 by wxsms
@jingyu-ml
2
quantification for SD1.5
#99 opened Oct 30, 2024 by zeng121
@jingyu-ml
1
Can we use multi GPU while exporting (diffusers ) onnx model?
#96 opened Oct 29, 2024 by wxsms
@jingyu-ml
5
cache_diffusion
#95 opened Oct 29, 2024 by zeng121
@jingyu-ml
cross_attention_kwargs ['adapter_params'] are not expected by DefaultAttnProcessor2_0 and will be ignored.
#94 opened Oct 25, 2024 by zeng121
2
cross_attention_kwargs ['adapter_params'] are not expected by DefaultAttnProcessor2_0 and will be ignored.
#93 opened Oct 23, 2024 by zeng121
@jingyu-ml
1
The algorithm output is not the same as after cache_diffusion
#92 opened Oct 23, 2024 by zeng121
@jingyu-ml
2
calib_dataloader created fail
#91 opened Oct 22, 2024 by relaxtheo
2
unrecognized arguments: --deployment
#90 opened Oct 22, 2024 by relaxtheo
@cjluo-omniml
1
quantization flux-dev int8 or fp8 occur oom in L40s
#89 opened Oct 21, 2024 by algorithmconquer
@jingyu-ml
1
Can you quantify the Multi-scale Deformable Attention module?
#88 opened Oct 12, 2024 by IEIAuto
ARM64 compatibility
#86 opened Oct 11, 2024 by felixkarevo
2
ProTip! Follow long discussions with comments:>50.