search

lucidrains / CoCa-pytorch

Implementation of CoCa, Contrastive Captioners are Image-Text Foundation Models, in Pytorch
MIT License
1.04k stars 88 forks source link

How to train the model using my own dataset? #9

Closed keepcodeandsmile closed 2 years ago

keepcodeandsmile commented 2 years ago

Can someone tell me how to train the model using my own dataset? is it like below？But I have many images and texts...

# train by giving CoCa your text and images with `return_loss = True`
loss = coca(
    text = text,
    images = images,
    return_loss = True  # set this to True to get the full caption + contrastive loss
)
rom1504 commented 2 years ago

You could check openclip training scripts as an example.

You need to train this model for several thousand GPU hours to get a small model trained.

On Tue, Aug 23, 2022, 13:52 keepcodeandsmile @.***> wrote:

Can someone tell me how to train the model using my own dataset? is it like below？But I have many images and texts...

train by giving CoCa your text and images with return_loss = True

loss = coca(

text = text,

images = images,

return_loss = True  # set this to True to get the full caption + contrastive loss

)

— Reply to this email directly, view it on GitHub https://github.com/lucidrains/CoCa-pytorch/issues/9, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAR437QJIVOC5V5TH64DNTLV2S3P3ANCNFSM57LFNPLQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>