Why is the Pre-trained Models of coco_caption empty in "VinVL_DOWNLOAD.md" file?

[Ammexm]

Hi~ Link: https://github.com/microsoft/Oscar/blob/master/VinVL_DOWNLOAD.md

Excuse me, why is the Pre-trained Models of coco_caption empty in "VinVL_DOWNLOAD.md" file? The link to the file is shown above. Has the download link of the model been transferred to another place?

The question needs to be answered urgently. Thank you very much.

[abhidipbhattacharyya]

Hi, Are you able to get the model? I am also trying to have the pre-trained caption model.

I can not download the feature files for flickr30k. The link seems to be broken.

[abhidipbhattacharyya]

@Ammexm I think you can download the models from here https://github.com/microsoft/Oscar/blob/master/VinVL_MODEL_ZOO.md#Image-Captioning-on-COCO

[Ammexm]

@Ammexm I think you can download the models from here https://github.com/microsoft/Oscar/blob/master/VinVL_MODEL_ZOO.md#Image-Captioning-on-COCO

Thx. for your answer! Do you mean the model is inside the Checkpoint package? I'll download it and check it out. Thx. again

[abhidipbhattacharyya]

@Ammexm I have not checked that yet but that is my best guess. I am still coding to make it work on my data. If you have already checked please let me know.

[Ammexm]

@abhidipbhattacharyya hi~really thx for you answering. I have been downloaded such checkpoints and I found it could be tested successfully on COCO test set. But I have not try to fine-tune it and I have no idea if it would be work as expected. It seems worthy to have a try.