search

HelloRicky123 / Siamese-RPN

Full reimplementation of siamese rpn, has 0.24 eao on vot2017.
MIT License
223 stars 44 forks source link

train the 54vision without yt_bb #30

Closed ghost closed 5 years ago

ghost commented 5 years ago

Hi, I trained your 54vision with VID and got Nan loss, then I reduce the LR it can run properly. I run the test_OTB.py with my trained model but error occured to data overflow, why？Awaiting your reply.

HelloRicky123 commented 5 years ago

Maybe the predicted box has w or h under 0, or has cx,cy out of image.

ghost commented 5 years ago

Thanks for your reply, I noticed your 54vision because it need't to train on yt_bb, which takes a lot of time, when I run your train codes without any change, loss increasing until Nan occurs. I wonder why this happens, is your codes' problem?

HelloRicky123 commented 5 years ago

I forget whether it use grad clip, if not, it should be tried several times to train.------------------ Original ------------------From: NatureRiver notifications@github.comDate: Fri,Apr 26,2019 4:34 PMTo: HelloRicky123/Siamese-RPN Siamese-RPN@noreply.github.comCc: HelloRicky123 qftju12345@163.com, Comment comment@noreply.github.comSubject: Re: [HelloRicky123/Siamese-RPN] train the 54vision without yt_bb(#30)Thanks for your reply, I noticed your 54vision because it need't to train on yt_bb, which takes a lot of time, when I run your train codes without any change, loss increasing until Nan occurs. I wonder why this happens, is your codes' problem?

—You are receiving this because you commented.Reply to this email directly, view it on GitHub, or mute the thread. {"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/HelloRicky123/Siamese-RPN","title":"HelloRicky123/Siamese-RPN","subtitle":"GitHub repository","main_image_url":"https://github.githubassets.com/images/email/message_cards/header.png","avatar_image_url":"https://github.githubassets.com/images/email/message_cards/avatar.png","action":{"name":"Open in GitHub","url":"https://github.com/HelloRicky123/Siamese-RPN"}},"updates":{"snippets":[{"icon":"PERSON","message":"@longqiwf in #30: Thanks for your reply, I noticed your 54vision because it need't to train on yt_bb, which takes a lot of time, when I run your train codes without any change, loss increasing until Nan occurs. I wonder why this happens, is your codes' problem?"}],"action":{"name":"View Issue","url":"https://github.com/HelloRicky123/Siamese-RPN/issues/30#issuecomment-486975831"}}} [ { "@context": "http://schema.org", "@type": "EmailMessage", "potentialAction": { "@type": "ViewAction", "target": "https://github.com/HelloRicky123/Siamese-RPN/issues/30#issuecomment-486975831", "url": "https://github.com/HelloRicky123/Siamese-RPN/issues/30#issuecomment-486975831", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { "@type": "Organization", "name": "GitHub", "url": "https://github.com" } } ]