ImperialNLP / VTLM

Cross-lingual Visual Pre-training for Multimodal Machine Translation

https://hucvl.github.io/VTLM

ImperialNLP/VTLM Issues

download data can' t be found
Updated 2 years ago
feature-extractor.py can't use
Updated 2 years ago
No train.order Files
Updated 2 years ago1
the process image file(.pkl)size
Updated 3 years ago
downloading from zenodo.org is too slow
Closed 3 years ago
Fix machine translation finetuning for image-based model
Closed 3 years ago
Multi30k Finetuning and Testing
Closed 3 years ago
Multi30K Data Preparation
Closed 4 years ago
Ignore sentences don't have corresponding image features
Closed 4 years ago
KL Divergence Implementation
Closed 4 years ago1
Train image-blind XLM model with Conceptual Captions en-de set
Closed 4 years ago1