Skip to content

Commit

Permalink
update readme
Browse files Browse the repository at this point in the history
  • Loading branch information
Deanplayerljx committed Oct 28, 2019
1 parent 2c11d3f commit 17019ce
Show file tree
Hide file tree
Showing 4 changed files with 4 additions and 656 deletions.
Empty file added data/.Rhistory
Empty file.
232 changes: 0 additions & 232 deletions dataloaders/README.html

This file was deleted.

11 changes: 4 additions & 7 deletions dataloaders/README.md
@@ -1,12 +1,9 @@
# Dean's note
`vcr.py` is the dataloader that loads the train & val sets for full models
## Note
`vcr.py` is the dataloader for base models.

`my_vcr.py` is the dataloader that loads the train & val sets for models without object detections.
`vcr_attribute.py` is the dataloader for model using attribute information.

`my_vcr_no_image` is the dataloader that loads the train & val sets for no image models
`vcr_attribute_new_tag.py` is the dataloader for model using both attribute information and new tags.

`my_vcr_det.py` is the dataloader that loads the train & val sets for models with detections.

`my_vcr_vgg.py` is the dataloader that loads the train & val sets for vgg net, the only difference is that it resize the input image to 224 \* 224 instead of 786 \* 384 (the size original author use).

Please choose them approriately according to your model needs, and import it in `models/my_train.py`.
417 changes: 0 additions & 417 deletions similarity_match/README.html

This file was deleted.

0 comments on commit 17019ce

Please sign in to comment.