Multi-Task-Transformer

Multi-Task-Transformer copied to clipboard

Where is the code of paper TaskExpert: Dynamically Assembling Multi-Task Representations with Memorial Mixture-of-Experts

6

hi Have you provided the code of paper TaskExpert: Dynamically Assembling Multi-Task Representations with Memorial Mixture-of-Experts?

hongge831

from . import iou3d_cuda ImportErrorImportError: : cannot import name 'iou3d_cuda' from 'detection_toolbox.iou3d'

IkeFwd

- [X] Downloaded pretrained model for cityscapes-3d - [X] installed requirements as mentioned here https://github.com/prismformore/Multi-Task-Transformer/pull/23 - while running inference script ``` !CUDA_VISIBLE_DEVICES=0 python inference.py --config_path='./configs/cityscapes3d/cs_swinB_taskprompter.yml' --image_path='/content/IMG_20200209_2301.jpg' --ckp_path='/content/drive/MyDrive/YOLOv8/TaskPrompter_CS_swinB_v2.pth.tar' --save_dir='/content/' ``` error...

op1009

TypeError: TaskPrompter.__init__() got an unexpected keyword argument 'default_cfg'

2

Steps done: 1. Clone repo 2. Download `.pth.tar` files 3. Run below commands ``` CUDA_VISIBLE_DEVICES=0 !python3 inference.py --config_path=configs/pascal/pascal_vitLp16_taskprompter.yml --image_path=/content/Screenshot7.png --ckp_path=/content/Multi-Task-Transformer/TaskPrompter/InvPT_pascal_vitLp16.pth.tar --save_dir=output ``` **Error** ``` Traceback (most recent call last): File...

op1009

kailashahirwar

Traceback (most recent call last): File "main.py", line 176, in main() File "main.py", line 152, in main end_signal, iter_count = train_phase(p, args, train_dataloader, test_dataloader, model, criterion, optimizer, scheduler, epoch, tb_writer_train,...

HereNowL

Thank u for your great work! Could you provide the config of InvPT base on ViT-B on the Pascal-Context dataset? Thanks!!!

Rook1ex

Channel Task Prompts Wrong

1

Hello Authors, In the Channel Task Prompt Learning Part (code: Multi-Task-Transformer/TaskPrompter/models/transformers/taskprompter.py L249), the line: chan_prompts = self.chan_proj(chan_x) has been comment, which means skipping the channel task prompt results when updates...

AaronCIH

Hi there, First off, I want to say your work on InvPT++ is really impressive—it’s clear a lot of thought and effort went into it! I’ve been exploring the project...

ierroric

Metric of edge detection

1

I am a beginner in multi-task learning and have a silly question: Why does the code use loss as the metric to evaluate edge detection, while the paper uses ODS-F?...

Wyt-ong