yixuan-qiao

Larger model

4

Awesome work, thanks for releasing! Is there some plans to further release larger models, such as BLIP-large or BLIP-xxlarge?

vqa tasks

1

The work on the VQA dataset is amazing. I would like to know when it will be released if you don't mind. I can't wait to give it a try.

pre-trained model

1

hi, the uniter-large.pt released is trained on in-domain & out-of-domain data or only in-domain data?

load T5 pretrained model from huggingface or google

8

We test the pre-training phase of T5 using model parallel and data parallel. It's really a easy framework to use with high GPU utilization. But there are some points that...

Hi, I'm curious about the input template you use when generating the queries in V2. In V1, i found it in [convert_msmarco_doc_to_t5_format.py](https://github.com/castorini/docTTTTTquery/blob/master/convert_msmarco_doc_to_t5_format.py) > segment = doc_title + ' ' +...

monoT5 fine-tuning process

25

Is there any plan to make the fine-tuning process of monoT5 & duoT5 public? We follow the steps in the paper to finetune the T5-base & T5-3B in pytorch framework...

Document segmentation

6

Before segment each document into passages by applying a sliding window of 10 sentences with a stride of five, may be some extra pre-processing method using regular expression? We simply...