TPT-48

VDI on TPT-48

If you have any questions, feel free to pose an issue or send an email to [email protected]. We are always happy to receive feedback!

The code for VDI is developed based on CIDA. CIDA also provides many baseline implementations (e.g., DANN, MDD), which we used for performance comparasion in our paper. Please refer to its code for details.

We approximate the Earth Moving distance with Sinkhorn Loss to accelerate computation. This is done by using the GeomLoss package.

W $\rightarrow$ E

How to Train for task W $\rightarrow$ E on TPT-48

python main.py -c config_TPT_48_WE (or)
python main.py --config config_TPT_48_WE

How to Use the Pretrained Model to Do Inference

1. Download the weight from here and unzip under the folder "pretrain_weight".
2. Run the following code:

    python inference.py -c config_TPT_48_WE_inference (or)
    python inference.py --config config_TPT_48_WE_inference

Both training/inference will produce the result in "result_save" folder, and you can use the result to do visualization.

N $\rightarrow$ S

How to Train for task N $\rightarrow$ S on TPT-48

python main.py -c config_TPT_48_NS (or)
python main.py --config config_TPT_48_NS

How to Use the Pretrained Model to Do Inference

1. Download the weight from here and unzip under the folder "pretrain_weight".
2. Run the following code:

    python inference.py -c config_TPT_48_NS_inference (or)
    python inference.py --config config_TPT_48_NS_inference

Both training/inference will produce the result in "result_save" folder, and you can use the result to do visualization.

Visualization of TPT-48's Domain Indices

1. Train/inference the VDI on TPT-48 with either "W $\rightarrow$ E" or "N $\rightarrow$ S" task.
2. Check your result in "result_save" folder, and then change the first 2 lines in "visualize_tpt_48_indices.py":

dates = "2023-03-11" # filling your own dates for experiments
time = ["12","56","54"] # filling the time for experiments. format: hour, miniute, second

3. Run the following code:

python visualize_tpt_48_indices.py

Your plot should be in the folder that saves the results of your TPT-48 experiment ("result_save/dates-time/visualization"). It will look similar as follows:

Loss Visualization during Training

We use visdom to visualize. We assume the code is run on a remote gpu machine.

Change Configurations

Find the config in "config" folder. Choose the config you need and Set "opt.use_visdom" to "True".

Start a Visdom Server on Your Machine

python -m visdom.server -p 2000

Now connect your computer with the gpu server and forward the port 2000 to your local computer. You can now go to: http://localhost:2000 (Your local address) to see the visualization during training.