A repository to explore the latest developments in LLMs and run experiments!
retrieval-augmented-generation.py
Experiments with a Retrieval-Augmented Generation pipeline using the LangChain and the HuggingFace frameworks.
generate_synthetic_data.py
Generates artifical data from the input content prompt provided using llamafiles version of TinyLlama-1.1B model.
Example of Input:
name,region
Rahul,India
Mark,America
Schneider,Germany
OKenny,Ireland
Mieke,Germany
The output that is generated is
1. Name: Rahul
Region: India
2. Name: Mark
Region: America
3. Name: Schneider
Region: Germany
4. Name: OKenny
Region: Ireland
5. Name: Mikee
Region: Germany
6. Name: Miek
Region: Germany
7. Name: Jake
Region: United States
8. Name: Sarah
Region: United Kingdom
9. Name: John
Region: United Kingdom
10. Name: Tom
Region: United Kingdom
model_fetcher.py
Fetches an model from the the HuggingFace library and runs an inference through it