j-min / CLIP-Caption-Reward

PyTorch code for "Fine-grained Image Captioning with CLIP Reward" (Findings of NAACL 2022)

https://arxiv.org/abs/2205.13115

j-min/CLIP-Caption-Reward Issues

onnx version ?
Updated 7 months ago
Select part of the data set
Updated a year ago
pre-training model
Updated a year ago4
Confusion about the config
Updated a year ago3
Factor in the weigthed reward sum is inconsistent with the paper
Closed a year ago2
phase 2 of clips
Closed 2 years ago2
Config file and performance reproduce
Updated 2 years ago11
About language_evaluation
Updated 2 years ago1
Phase 1 validation throws: shape '[4, -1, 512]' is invalid for input of size 64000
Updated 2 years ago3
Where is the file 'data/ finecapeval. json'?
Closed 2 years ago1
Error running python 'clip_prepro_feats.py'
Closed 2 years ago2
difference between clip_prepo_feat.py and clipscore_prepo_feat.py
Closed 2 years ago2