Drama LLM - Project Ollama is a TypeScript and React-based application designed to interact with machine learning models for generating text-based responses. It provides a user-friendly interface for selecting different models and customizing parameters such as temperature, top-p, and top-k to fine-tune the generation process. Additionally, it allows users to define roles and input content for the models to process.
- Model Selection: Users can choose from a variety of pre-defined models to generate text.
- Customizable Parameters: Fine-tune the model's behavior by adjusting settings like temperature, top-p, and top-k.
- Role-Based Interaction: Define the role (system, user, assistant) for different types of interactions.
- Responsive Design: A user-friendly interface that adapts to various screen sizes.
To set up the project locally, follow these steps:
- Clone the repository:
git clone https://github.com/rohit1901/drama-llm.git
- Navigate to the project directory:
cd drama-llm - Install dependencies:
npm install
- Start the development server:
npm run dev
To configure the application, create a .env file in the root directory of the project with the following variables:
DRAMA_LLM_SECURITY: Set this todisableto bypass security checks during development. Note: This should only be used in a development environment. Default isdisable.DRAMA_LLM_ETH_LOGIN: Set this todisableto enable Ethereum login and SIWE (Sign-In with Ethereum) checks. The application will crash if no wallet is found. Default isdisable.
Example .env file:
DRAMA_LLM_SECURITY=disable
DRAMA_LLM_ETH_LOGIN=disableAfter starting the application, navigate to http://localhost:5173/ in your web browser. Select a model from the dropdown menu, adjust the desired parameters, and input your content. Press the "Generate" button to see the model's output.
Contributions are welcome! Please open an issue or submit a pull request for any bugs, features, or improvements.
This project is licensed under the MIT License - see the LICENSE file for details.