module.load_state_dict

[natank1]

What is the correct way to use model.load_state_dic (A model that was trained in GPU and may can be used for inference in CPU)?

[henon]

I don't know exactly. A unit test needs to be written for that method, it is implemented but currently still untested.

[natank1]

So what is the way to upload a pre trained torch model ( trained by GPU) to csharp? I wish to develop an inference engine in csharp

Thanks

נשלח מה-iPhone שלי

‫ב-18 באוג׳ 2019, בשעה 19:48, ‏‏Meinrad Recheis ‏notifications@github.com כתב/ה:‬

I don't know exactly. A unit test needs to be written for that method, it is implemented but currently still untested.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[henon]

I guess you need to read it up in the pytorch docs. let me know if anything you need for that is missing

[natank1]

sure

On Sun, Aug 18, 2019 at 9:23 PM Meinrad Recheis notifications@github.com wrote:

I guess you need to read it up in the pytorch docs. let me know if anything you need for that is missing

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SciSharp/Torch.NET/issues/5?email_source=notifications&email_token=AB7W2ZYNJAQHTEJ72YLF76TQFGHSNA5CNFSM4IMO6TZKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4RFLPQ#issuecomment-522343870, or mute the thread https://github.com/notifications/unsubscribe-auth/AB7W2Z5YJTYO3RMWFTA5E2TQFGHSNANCNFSM4IMO6TZA .