search

FactoDeepLearning / MultitaskVLFM

MIT License
25 stars 1 forks source link

Request the release of your pretrain model #1

Open YCaigogogo opened 1 year ago

YCaigogogo commented 1 year ago

Dear developers, Hello, I would like to ask if you can open the download path of the trained model used in your paper, thank you very much!

FactoDeepLearning commented 1 year ago

Here it is: https://zenodo.org/record/8124014

YCaigogogo commented 1 year ago

May I ask whether the text encoder of clip in this pre-trained model has been fine-tuned by the method in the paper?

FactoDeepLearning commented 1 year ago

Yes it has been fine-tuned (not all layers but only the last transformer layer + the projection layers)

YCaigogogo commented 1 year ago

Thank you for your reply. And I've got what might be a simple problem. I found this error message when I try to run "python3 Multitasking/main.py --load-pretrain=True --train=False". And My current path is "/data/yic/MultitaskVLFM-main". It is strange. How can I solve it.

================================================================= Traceback (most recent call last): File "/data/yic/MultitaskVLFM-main/Multitasking/main.py", line 4, in from Multitasking.dataset_utils.cub200_dataset import CUBDataset, check_or_download_CUB200_dataset, check_or_download_CUB200_oracle_seg ModuleNotFoundError: No module named 'Multitasking'

FactoDeepLearning commented 1 year ago

As stated in the readme, "pip install -e ." Should solve this issue