fairseq-image-captioning
fairseq-image-captioning copied to clipboard
krasserm
Transformer-based image captioning extension for pytorch/fairseq
Hey there, I want to generate captions on my own dataset. I added my images in the required folders but Im not able to generate the captions. Its says the...
In preprocess/tokenize_captions.py, ``` def load_annotations(coco_dir): with open(os.path.join(coco_dir, 'annotations', f'captions_train2014.json')) as f: annotations = json.load(f)['annotations'] with open(os.path.join(coco_dir, 'annotations', f'captions_val2014.json')) as f: annotations.extend(json.load(f)['annotations']) return annotations ``` It seems that this code was...
I am trying to run the demo file but it is saying metadata.csv not found
karpathy_test_resnet101_faster_rcnn_genome.tsv karpathy_train_resnet101_faster_rcnn_genome.tsv.0 karpathy_train_resnet101_faster_rcnn_genome.tsv.1 karpathy_val_resnet101_faster_rcnn_genome.tsv I try download with download link, but it's fail. The file is too big to download fail. I'm trying use "--features grid", but I don't know...
In wip-train-inception branch, the checkpoint_demo.pt file link directs to error page
Hi, Krasser When I run the demo script in colab I got the following error. Could you help me with this issue? I have installed all the packages required. Thanks...
I have my own dataset of (image, caption) pairs on which I'd like to train the model. Does this repository enables to do that without first extracting features/bounding boxes? Can...
Hi, I encountered some errors during the Self-critical sequence training stage: WARNING: attempting to recover from OOM in forward/backward pass Is this because the GPU memory is not enough? It...
Hi, I wonder if we can use the extracted features from the resnet-152 model, not from the Faster-RCNN because the former is easy to implement.
Following up on https://github.com/pytorch/fairseq/issues/759#issuecomment-589498214, it would be great if Faster-RCNN could be used directly, so we could input images instead of pre-computed features from MS-COCO. Regarding the specific Faster-RCNN PyTorch...
