Zhuosheng Zhang comments

Results 24 comments of


                                            Zhuosheng Zhang

Reproduction results for SQuAD 2.0

@rakshanda22 Unfortunately not yet :(

Cannot re-produce the result on SNLI

Sorry for a late reply. This might be because of the version mismatch as the repo has continuously updated after the model release. Could you replace the L1024 with L1024...

I am not able to run the model for inference. I am trying to run model1 from codalab using sh.albert script but the error says the model is corrupted. Please help

Hi, our submitted single model is "model4". Please use the models in the dependencies if you still need it. https://worksheets.codalab.org/bundles/0x9ca59fbdc85d4578a2c485c936c7dbc5

click accuracy and scroll accuracy

Hi, please refer to the following codes. In Section 5.1, the text accuracy is only measured if the predicted and reference texts are matched or overlapped (see below). It is...

Link for Dataset and trained models is not working.

Sorry that I forgot to make it public. It should work now :)

Link for Dataset and trained models is not working.

Auto-UI-base, Auto-UI-large represent the base- and large- size models. Blip is the preprocessed data (the screenshots have been converted into vision features). If you want to see the original data,...

Link for Dataset and trained models is not working.

Hi @Yangyi-Chen, Thanks for your interest. You can refer to https://github.com/cooelf/Auto-UI/tree/main/dataset for the data split. The files contain the episode ID that you can use to fetch the original dataset.

在提供链接的blip数据集并未找到用于inference的single_parsed_episode_t5_blip数据

The inconsistent name has been fixed. Please try the updated version.

I can't find main_central.py.

Hi. It is a typo. Please change it to main.py

[17:28:39] [Model]: Loading declare-lab/flan-alpaca-large...

Hi, did you try to conduct a unit test to see if it is possible to load a pre-trained model using huggingface? My guess is that the memory is not...