In the first Course, you will develop the core translation pipeline using Haystack's NLP components with an underlying LLM. You will implement the translation routine along with dedicated routines to preprocess incoming code and postprocess the translated output, setting a strong foundation for further enhancements.
Overview
Syllabus
- Unit 1: Exploring Haystack Basics
- Connecting Your First Haystack Pipeline
- Adding Components to Your Pipeline
- Supplying Input for Your Pipeline
- Switching Your Pipeline to Spanish
- Switching to a New Language Model
- Build a Complete Translation Pipeline
- Unit 2: Building a Basic Code Translation Pipeline
- Perfecting the Code Translation Pipeline
- Connecting Your Code Translation Pipeline
- Mastering Prompt Templates for Translation
- Connecting Your Code Translation Pipeline
- Build a Complete Code Translator
- Unit 3: Implementing Preprocessing Routines for Code Translation
- Powering Up the Preprocessor
- Declaring Outputs for Your Preprocessor
- Detecting the Language of Code
- Refining Code Extraction with Prompts
- Build a Complete Code Preprocessor
- Unit 4: Implementing Postprocessing Routines in Code Translation
- Prompting for Clear Code Explanations
- Making Postprocessing More Reliable
- Adding Flexible Explanation Levels
- Build a Complete Code Postprocessor
- Unit 5: Adding Guardrails to Prevent Exploitation
- Guardrails That Never Miss
- Connecting Guardrails for Safe Translation
- Clear Instructions for Reliable Guardrails
- Routing Decisions for Safe Translation
- Build a Secure Code Translation Pipeline
- Adding Explanations to Guardrails