- Save History Toggle: Implemented a
save_historyswitch inconfig.yamlto control chat history saving.- Default value
falsefor new configurations. save_historystatus is read fromconfig.yamlon startup and when switching configurations.- Validation ensures
save_historycan only betrueorfalse.
- Default value
- Conditional Chat History Saving: Chat history is now saved only when
save_historyis enabled in the configuration.
- Updated Menu Format: The main menu now displays as:
The
[i]nfo | [m]odel | [u]pdate | [h]istory ([s]ave:On/Off) | [c]onfig | [q]uit[s]avestatus dynamically showsOnorOffbased on thesave_historysetting. - History Toggle Command: The command to toggle history saving is
s. - App Title Display: The
app_titlefromconfig.yamlis now consistently displayed before the menu in all relevant instances:- On application startup.
- After toggling history saving.
- After changing configuration files.
- After updating models.
- Response Statistics: Added detailed statistics after each response showing:
- Total tokens used
- Completion time in seconds
- Tokens per second processing rate
- Multi-Provider Support: Now compatible with any OpenAI-compatible API provider
- Works with xAI's grok-3-mini-beta and other compatible models
- Simply configure the base_url in your config.yaml
- Enhanced Error Handling: Better error messages when API responses are empty or malformed
- Response Processing: Improved handling of API responses with proper validation
- Documentation: Updated README with new features and usage examples
- Fixed issue with response not being displayed in the terminal
- Fixed timing calculation for completion time when not provided by the API
- Removed debug output that was appearing in the terminal
- Refactored the Chat method in the client.go file to properly handle API responses
- Added validation to ensure responses contain valid content
- Implemented fallback timing calculation when the API doesn't provide completion time
- fixes: improve UI formatting and readability
- Minimalistic CLI interface for interacting with Groq AI models
- One-shot prompts with no conversation context
- Interactive commands: [i], [m], [h], [q]
- Model selection and information viewing
- Chat history saved as Markdown files
- Configuration via YAML and environment variables
- Cross-platform support (Windows, Linux, macOS)
- Docker images (Debian and RHEL-based)
- Low resource usage (11-14MB RAM)
- Small Docker image size (9-11MB)