Issues
- 1
retrieval only mode
#36 opened by oferidan1 - 2
Can I use the embedding for training
#35 opened by LiJichen0114 - 0
Evaluation code of VQAv2
#34 opened by Yui010206 - 1
- 2
- 0
=> no checkpoint found at '=/home/...
#32 opened by eveningwalk - 7
- 2
I got Unexpected key(s) in state_dict error
#30 opened by eveningwalk - 5
- 2
What is CC3M Embeddings
#27 opened by ziqipang - 3
[RET] Embedding
#26 opened by pUmpKin-Co - 4
Dealing with Corrupted Images in CC3M
#25 opened by ziqipang - 0
The ability of in-context learning
#24 opened by yongliang-wu - 4
Evaluation code for VQAv2
#23 opened by ys-zong - 7
Computing output likelihoods with the model
#13 opened by vishaal27 - 2
Huggingface pipeline
#15 opened by Marcusntnu - 6
The reproduction of FROMAGe training
#22 opened by Ziyang412 - 2
The cross entropy loss in training stage
#21 opened by Ziyang412 - 8
The evaluation speed of IT2T on VisDial
#20 opened by Ziyang412 - 1
Evaluation for VisDial
#19 opened by Ziyang412 - 3
- 5
- 3
Choice of retrieval embedding dimension q = 256
#10 opened by EIFY - 2
How does generate work?
#16 opened by zhaoshitian - 5
- 3
What is "fromage_vis4" model?
#8 opened by ahnjaewoo - 2
How to load dataset?
#12 opened by zhaoshitian - 1
Failure in testing the demo
#9 opened by Yingjia-Wan - 2
- 2
Question about the frozen language model
#6 opened by sijeh - 2
Should the last_embedding_idx = caption - 2 ?
#4 opened by sijeh - 6
- 3
when the source codes can be released?
#1 opened by runzeer - 1
Asking for roadmap with more details?
#2 opened by ZeinabTaghavi