Skip to content

Latest commit

 

History

History
56 lines (40 loc) · 1.76 KB

TRAINING_INSTRUCTIONS.md

File metadata and controls

56 lines (40 loc) · 1.76 KB
Raw

Training Instructions for Vishwamai Model

Prerequisites

  1. Python Environment: Ensure you have Python 3.8 or later installed.
  2. Virtual Environment: It is recommended to use a virtual environment to manage dependencies.
  3. Git LFS: Ensure Git LFS is installed and initialized in the repository.

Setup

  1. Clone the Repository:

    git clone https://github.com/VishwamAI/chat-agent.git
cd chat-agent

  2. Initialize Git LFS:

    git lfs install
git lfs pull

  3. Create and Activate Virtual Environment:

    python3 -m venv venv
source venv/bin/activate

  4. Install Dependencies:

    pip install -r requirements.txt

Dataset Preparation

Ensure the dataset is available in the datasets directory. The config_for_9b.yaml file is configured to use datasets/dev.json for both training and validation.

Training the Model

  1. Run the Training Script: 
    python scripts/train_t5.py --config configs/config_for_9b.yaml

Notes

  • The train_t5.py script is designed to train the Vishwamai model using the specified configuration file.
  • Ensure that the datasets/dev.json file is correctly formatted and available in the datasets directory.
  • The training process may take a significant amount of time, depending on the size of the dataset and the available computational resources.

Troubleshooting

  • If you encounter any issues with missing dependencies, ensure that all required packages are listed in the requirements.txt file and installed in your virtual environment.
  • For any other issues, refer to the repository's README file or seek assistance from the repository maintainers.

This PR was written by Devin 👼