how to create char CNN-RNN text embedding for our dataset please?!!!!

[Maymaher]

I'm trying to run StackGAN on my dataset but I need to create char cnn-rnn text embedding for my dataset
any help??!

[Ereebay]

Follow the steps of this repo
https://github.com/reedscot/cvpr2016

[saramahersalem]

@Maymaher
I have the same problem , did you get it ?

[Maymaher]

No , but I use AttnGANs model, it creates embedding from scratch ‫في الجمعة، 12 يوليو 2019 في 4:58 م تمت كتابة ما يلي بواسطة ‪Sara Maher‬‏ <‪ notifications@github.com‬‏>:‬

…

@Maymaher <https://github.com/Maymaher> I have the same problem , did you get it ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#58?email_source=notifications&email_token=AIIVP4FM4AAFKBV56PSRH5DP7CL2LA5CNFSM4HZFAZMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZ2AARA#issuecomment-510918724>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIIVP4CHLYGLGQHE27V5KATP7CL2LANCNFSM4HZFAZMA> .

[Maymaher]

i faced problem in this also, the birds and flowers dataset are in a torch format and I can't convert my data same as this format ‫في الأربعاء، 10 يوليو 2019 في 6:19 م تمت كتابة ما يلي بواسطة ‪Xinkang Jia‬‏ <‪notifications@github.com‬‏>:‬

…

Follow the steps of this repo https://github.com/reedscot/cvpr2016 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#58?email_source=notifications&email_token=AIIVP4GJLXLR5225BBXOXY3P6YD2FA5CNFSM4HZFAZMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZT7WUY#issuecomment-510131027>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AIIVP4D6RNXHCGVK4NI5GFTP6YD2FANCNFSM4HZFAZMA> .

[jszgz]

Does anyone know what does the second dimension in embeddings represent? Why do we pick a randint embedding_ix in the second dimension?
embedding_ix = random.randint(0, embeddings.shape[0] - 1) embedding = embeddings[embedding_ix, :] if self.target_transform is not None: embedding = self.target_transform(embedding) return imgs, wrong_imgs, embedding, key

[weili-git]

i faced problem in this also, the birds and flowers dataset are in a torch format and I can't convert my data same as this format ‫في الأربعاء، 10 يوليو 2019 في 6:19 م تمت كتابة ما يلي بواسطة ‪Xinkang Jia‬‏ <‪notifications@github.com‬‏>:‬
…
Follow the steps of this repo https://github.com/reedscot/cvpr2016 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#58?email_source=notifications&email_token=AIIVP4GJLXLR5225BBXOXY3P6YD2FA5CNFSM4HZFAZMKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZT7WUY#issuecomment-510131027>, or mute the thread https://github.com/notifications/unsubscribe-auth/AIIVP4D6RNXHCGVK4NI5GFTP6YD2FANCNFSM4HZFAZMA .

have you fixed the problem about converting images and captions into t7 format just like in cvpr2016_flowers?

[rishi-wqd190004]

Is there any update on this to create char-CNN-RNN model for your own dataset?