VLSI HealingSPA

Setting Up the Environment

To set up the environment, use the provided vlsi_env.yaml Conda configuration file. Run the following command:

conda env create -f vlsi_env.yaml

Running the Application

Single GPU Environment

If you are using a single GPU, simply run:

python main.py

Multi-GPU Environment

For multi-GPU setups, follow these steps:

1. Configure accelerate by running:

   accelerate config

   Refer to Accelerate Configuration Guide for detailed instructions.

2. Launch the training process:

   accelerate launch main.py

Running Inference

After training, you can run inference using the following command:

python inference.py --checkpoint <CHK_PATH>

Replace <CHK_PATH> with the path to your trained model checkpoint.