You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
---原始邮件---
发件人: "Dahan ***@***.***>
发送时间: 2023年8月27日(周日) 晚上9:54
收件人: ***@***.***>;
抄送: ***@***.******@***.***>;
主题: Re: [THU-MIG/torch-model-compression] qat量化能够导出onnx,能够在tensorrt上进行部署吗? (Issue #23)
原理和torch的qat是一致的,都是额外加onnx node来存储量化信息,所以理论上肯定能在 trt 跑,只是仓库里目前的代码不一定跑得通,可能需要你看着改改
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
No description provided.
The text was updated successfully, but these errors were encountered: