Zonglin DI

Results 30 comments of Zonglin DI

How to evaluate the image difference description?

BTW, I understand that the image difference description dataset will greatly hurt the benchmark's performance. Is my understanding correct? If my understanding is correct, what do you mean by the...

Can I use a single gpu to train this model?

Got it. Thanks a lot!

Can I use a single gpu to train this model?

Hi Bo, I tried the following script, but I got an error. I want to train the model using LA dataset and I added the `image_processor` and `train_num_samples` back to...

Can I use a single gpu to train this model?

Yes. My software versions are as follows. `transformers`: 4.30.2 `accelerate`: 0.20.3 `torch`: 2.0.1 Sure. I can try the latest branch. Thanks!

[Fix/Train] `image_processor` is missing in `get_mimicit_dataset`

Thanks!

Onedrive link seem gone

I am sorry that I did not clarify the specific link. The link which cannot be opened is [scene graph](https://onedrive.live.com/embed?cid=22376FFAD72C4B64&resid=22376FFAD72C4B64%21779871&authkey=AA33n7BRpB1xa3I). Could you please upload it to Google cloud? Thanks!

[FEATURE REQUEST] MMC4 Dataset Preprocess

Thanks for your update. I will follow up on Feature 2.

Can I train OpenFlamingo without LIAON dataset?

Thanks for your kind reply. I am planning to contribute a PR to make this project more complete :). I will close this issue after I finish the PR.

[rank2]: AttributeError: 'Collator' object has no attribute 'tokenizer'

https://github.com/ElegantLin/Mantis/pull/1

param error can not connect qwen-vl-max

Hi, do you find the solution?