I have trained the model on my own dataset, and got tensorflow checkpoint. In the readme, it says that we can use embed.py to compute the embeddings of the images. However, I want to input just a list of images and get embeddings to some file like embeddings.csv not h5 (like the way trinet_embed processes)?
Is that possible? I appreciate your attention.
I have trained the model on my own dataset, and got tensorflow checkpoint. In the readme, it says that we can use embed.py to compute the embeddings of the images. However, I want to input just a list of images and get embeddings to some file like embeddings.csv not h5 (like the way trinet_embed processes)?
Is that possible? I appreciate your attention.