Reproducible Builds
Step-by-step implementation guides with GitHub repositories, Dockerfiles, notebooks, local vector databases, environment setup, and troubleshooting.
Run Your First EmbeddingGemma 2 Search Pipeline Locally
This post details how to run a minimal, local semantic search pipeline using EmbeddingGemma 2 on ordinary hardware like ...
A Dockerized EmbeddingGemma 2 and Qdrant Workflow for Developers
This post transitions the previous local Python embedding pipeline into a robust two-container Docker Compose setup usin...
Build a Hybrid Retriever with BM25 and Dense Embeddings
This post extends the Dockerized EmbeddingGemma 2 and Qdrant stack by adding a BM25 index and a simple weighted hybrid f...
How to Turn a Notebook Experiment into a Repeatable Command
This post details how to transition from exploratory Jupyter notebooks to clean, repeatable command-line scripts. The au...
The Five Files Every Reproducible Retrieval Repository Should Include
This post concludes the Reproducible Builds series by identifying the five essential files every retrieval repository mu...