add lacking typing

Refactor stop sequences in chat.py for improved template handling

Add RAG generation (#19)

Add requirements.txt for synthetic dataset generator dependency, update import paths for InferenceEndpointsLLM in multiple files, and modify example deployment script to comment out HF_TOKEN.

Update version to 0.1.6, remove requirements.txt, and enhance dataset handling in pipelines. Added Gradio support and improved LLM class retrieval. Commented out HF_TOKEN in example deployment script.

add randomisation of system prompts for generation

update deployment with API providers

add qwen stop sequences

update sft to chat