configuration.md

Configuring telegoGPT

telegoGPT can be customized and configured to suit your preferences using the config.yaml file.

Configuration Options

token

• Description: Telegram Bot Token obtained from BotFather.
• Example: token: YOUR_TELEGRAM_BOT_TOKEN

max_response_length

• Description: Maximum length (in characters) of the generated response.
• Example: max_response_length: 500

temperature

• Description: Controls the randomness of the response. Lower values make the responses more focused and deterministic, while higher values make them more random.
• Example: temperature: 0.7

top_p

• Description: Controls the nucleus sampling. A higher value will limit the response to the most likely tokens based on a cumulative probability.
• Example: top_p: 0.85

frequency_penalty

• Description: Controls the penalty for using frequent tokens. A higher value will make the responses more focused.
• Example: frequency_penalty: 0.2

presence_penalty

• Description: Controls the penalty for using new tokens. A higher value will encourage the model to use a broader vocabulary.
• Example: presence_penalty: 0.0

stop_sequence

• Description: Sequence of characters that, if encountered, will signal the end of the response.
• Example: stop_sequence: "###"

Modifying the Configuration

1. Open the config.yaml file in the project root using a text editor.

2. Adjust the configuration options as per your preferences.

3. Save the file after making the necessary changes.

Applying Customizations

Restart the telegoGPT bot after modifying the configuration for the changes to take effect.

---

For more information, visit the telegoGPT GitHub repository.