Multimodal Hashtag Prediction with instagram data & pytorch (2nd Place on OpenResource Hackathon 2019)
Multimodal hashtag prediction from instagram
ALBERT
for text, VGG16
for image[CLS]
token from ALBERT, and change it to 100 dim$ docker run -d -p 80:80 adieujw/hashtag:latest
Copy the id. It will be used when you give request.
# URL
localhost:80/predict?image_id=1DGu9R5a9jpkY-fy79VrGFmCdJigzTMC-&text=20%20days%20till%20Christmas%20%F0%9F%98%8D%F0%9F%8E%85&max_seq_len=20&n_label=10
image_id
: the share id you can get from google drive abovetext
: like caption in instagrammax_seq_len
: maximum sequence lengthn_label
: num of labels you want to predicthttps://endpoint.ainize.ai/monologg/hashtag/predict?image_id={image_id}&text={text}&max_seq_len={max_seq_len}&n_label={n_label}
# URL
https://endpoint.ainize.ai/monologg/hashtag/predict?image_id=1DGu9R5a9jpkY-fy79VrGFmCdJigzTMC-&text=20%20days%20till%20Christmas%20%F0%9F%98%8D%F0%9F%8E%85&max_seq_len=20&n_label=10