WordSmith: Your Professional Correspondent

wordsmith.demo.mp4

Create Virtual environment

python -m venv my_hf_env

Activate Virtual environment

On macOS/Linux
source my_hf_env/bin/activate

On Windows
my_hf_env\Scripts\activate

After activation, your command line prompt will change to show the environment's name, like (my_hf_env)

install dependencies:

pip install -r requirements.txt

Note: If you have a GPU, read the commented lines.

Choose your dataset:

I have provided a few datasets to train the model on, but you can outsource a dataset as well. Make sure to replace the dataset value in train.py.

First off, run

python main.py
To initialize the project.

Train the model:

Once you have decided what you want to fine-tune the model on, and made changes in the train.py script accordingly,
Run:
python train.py
P.S. This is going to take a good while

Once the model is done training

You can interact with it by running:
python inference.py