Could you let me know how to run inference mode?

[DBpackage]

Hello! I've run your learning model.

The training and testing processes were very straightforward, making it easy to execute. Thank you for that.

However, I am experiencing some confusion as there is no information provided on how to perform inference on the test data using the saved ckpt file.

Could you possibly assist me with this running inference(or test) mode code?

Sincerely,

[whitehat32]

You can use the command below to test on a saved checkpoint file: python main.py --root_data_path your_dataset_path --ckpt_path your_ckpt_path --mode predict (I have updated main.py to support testing on saved checkpoints.)

[DBpackage]

Thanks for fast reply! I'll try it.

And I have one more question!

Could I apply classification on your model? If I can, may I know how to run the classification model? I've tried to do that with args --objective classification I made additional 'CLS_LABEL' column in my training/valid/test dataset but it didn't work.

Thanks in advance! sincerely

[whitehat32]

Do you mean the model is still training on the regression task?

[DBpackage]

Hi!

For the regression task, I confirmed that the model is working. However, I didn't know how to proceed with the prediction using the trained model, so I asked the first question. Now, it's clear! I confirmed that prediction mode with regression task is working!

But the other question about the classification task, I've tried to run classification mode with args --objective classification But it didn't work. I added new 'CLS_LABEL' column which has binary labels (0,1) in my train/valid/test dataset. However, I got systematic error with tensors

If you want to check the error, then check this file Traceback (most recent call last).txt

I'm trying to find the problem but it's not clear. I'm pretty newbee with pytorch lightning modules. Have you confirmed that the classification task working?

Sincerely