haiyang-w / git Goto Github PK

I've encountered the following warnings when trying fast-mode,
I've tried loading BlipTokenizer, but why turned out to be the "BlipTokenizer" warning?

The tokenizer class you load from this checkpoint is not the same type as the class this function is called from. It may result in unexpected tokenization.
The tokenizer class you load from this checkpoint is 'BertTokenizer'.
The class this function is called from is 'BlipTokenizer'.

Besides, my demo run on an offline env,
where I can not download pretrained weights online

Is there any suggestions? Many thanks!

I try to train the single detection task with
"bash tools/dist_train.sh configs/GiT/single_detection_base.py 1 --work-dir output/test"

Could i ask you how to solve the problem?

Thank you!

Hi,

Thanks for making the code public. I'm trying to run few shot learning on LoveDA dataset. Once I start training, I get the following error.

FileNotFoundError: [Errno 2] No such file or directory: 'data/loveDA/ann_dir/train/1040.png'

My loveDA folder structure is as follows:

GiT
|──data
│   ├── loveDA
│   │   ├── img_dir
│   │   │   ├── train
│   │   │   ├── val
│   │   │   ├── test
│   │   ├── ann_dir
│   │   │   ├── train
│   │   │   |      |----- 1040.png.json
│   │   │   ├── val

as described here: https://github.com/Haiyang-W/GiT/blob/main/tools/dataset_preprocess/dataset_prepare.md

However, I see in the same link that I should run python tools/dataset_converters/loveda.py /path/to/loveDA but I don't see the file in the cloned repo. If this code is needed to fix this issue, can you please provide me with the dataset_converter file for loveda dataset?

I try to use this awesome model like below

but it's not working.

Could i ask you what is the problem?

Thank you!

Hi,

Thank you for the code and the Readme which are both very well organized. I am trying to setup the few-shot and zero-shot datasets. Is there any details I need to take account of?

Thank you!.

The following line in git_det_head, skips the last token (height) for input tokens:
# input tokens for parallel training
input_tokens = targets_tokens[:, :-1]

This leads to seq_embed skipping height during training. Is this intended? If so, could you please clarify why this is done?

Thank you!.

There are discrepancies in the results of few-shot results and in one zero shot result as shown below:

However, results for zero-shot cs-det, cs-seg and sun - rgbd seg match with only slight deviations.

Would it be possible that I am looking at the wrong metric or maybe there is some other issue?
Would be glad if you could help me with this. Thank you!.

First, thanks for sharing the codes and models. I just wonder in this paper, how the COCO mAP metric was calculated under the next token prediction decoder output. Since traditionally, for computing mAP, we need a confidence score prediction for each box, but for next-token prediction output, how this score can be obtained? or am I missing something?

能否提供hugging face的demo试玩？

When using the large and huge few-shot configurations for loveda dataset, since the config uses both the loveda base config and the git_large config, there are overlapping keys. Issue can be resolved by creating a copy of loveda_base and replacing git_base with git_large and load_from base to large.

Is there any other way to fix this? Or am I missing something here?

Thank you!

demo 下面都是检测和分割的，没有提示词的