JJJYmmm

When I run /colab/pix2seq_inference_object_detection.ipynb, I meet an error as shown in the following figure. It shows that the problem lies in line233 and line243 in /tasks/visualization/vis_utils.py. ``` display_str_heights = [font.getsize(ds)[1]...

fix /tasks/visualization/vis_utils.py

1

fix bug in #48

## Fix variable naming errors `codebook_dim` -> `embedding_dim` `num_tokens` -> `n_embed`

Thanks for your video. Since the transformer take in the quantified image token generated by VQGAN, which codebook has indices (0~n_embed-1), and transformer’s sos token is also set to zero...

Hello, excuse me. Are this images in the LIP dataset cropped based on the bounding boxes annotated in MSCOCO, or you just annotate in manual?

It seems that the ViTVQ only consider the origion VQVAE(Not considering codes has 3 dimensions, e.g. HxWxN, N means residual times in RQVAE) https://github.com/thuanz123/enhancing-transformers/blob/1778fc497ea11ed2cef134404f99d4d6b921cda9/enhancing/modules/stage1/vitvqgan.py#L81 RQ-VAE is only mentioned here(and flag...

Update conversation.py

1

fix calculation about `shortest_edge` https://github.com/haotian-liu/LLaVA/issues/1413

Fix the conditional determination of whether the scores variable is empty. ## problem When running evaluate_refcoco.sh, encounter errors below. ```bash scores:tensor([1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,...

Question about data processing in pretrain data.

1

Why the following actions were taken. Is there anything special about cc12m I missed? https://github.com/OFA-Sys/OFA/blob/a36b91ce86ff105ac8d9e513aa88f42b85e33479/data/pretrain_data/unify_dataset.py#L321-L323 Looking forward to your reply.

some bugs about test.py

2

Hi Shariatnia, I meet some trouble in test.py ## issue 1 Firstly, function `postprocess` **does not consider the situation where the model detect nothing**,so it will pass empty sequence(\\) to...