krasserm/fairseq-image-captioning
Transformer-based image captioning extension for pytorch/fairseq
PythonApache-2.0
Issues
- 2
why loss is always 0.000 in epoch 1 ?
#33 opened by nvnhann - 8
- 0
Testing on own dataset
#32 opened by arsalananwar11 - 1
- 0
meta.csv not found
#30 opened by akshatsingh1718 - 0
how can I get these file?
#29 opened by andy9420 - 0
- 0
FileNotFound Error in demo
#27 opened by Haochen-Luo - 0
Training end-to-end on my own dataset
#26 opened by MiriamFarber - 2
@krasserm : Can you suggest how to use this repo to inference on our own test dataset?
#25 opened by Mas-Y - 10
Preprocess issue
#19 opened by vinod-13 - 5
teacher forcing ratio
#24 opened by goodluck110706112 - 10
- 1
FileNotFoundError: [Errno 2] File b'output/valid-features-obj/metadata.csv' does not exist: b'output/valid-features-obj/metadata.csv'
#22 opened by nakulp21 - 10
Performance reproduce
#20 opened by Beanocean - 2
- 2
Query regarding COCO dataset
#18 opened by Ashwin-Ramesh2607 - 4
What if using a resnet-152 when extracting features instead of the Faster-RCNN?
#17 opened by Kyubyong - 7
Use Faster-RCNN directly
#9 opened by adrelino - 3
- 3
Some bug in caption.py
#15 opened by hiteshK03 - 3
- 0
Train new model using custom data
#12 opened by AliOsm - 0
Bounding box encoding with sin/cos functions
#11 opened by krasserm - 0
- 0
Implement Meshed-Memory Transformer
#7 opened by krasserm - 1
Option to fine-tune InceptionV3 encoder
#4 opened by krasserm - 0