Field Notes
Our Blog
Insights on AI, Machine Learning, Web Development, and emerging technologies from industry experts.
// jump to
PristrenPristrenInsights on AI, Machine Learning, Web Development, and emerging technologies from industry experts.
// jump to
Built by Pristren
Reading about AI tools? Run your team on Zlyqor — chat, meetings, projects, and time tracking in one workspace.
241–252 of 523
How to build a retrieval-augmented generation system using Ollama, ChromaDB, and Sentence Transformers. When open source RAG beats paid options.
Mahmudul Haque Qudrati
CEO & ML Engineer
Hugging Face hosts 900k+ models, datasets, and Spaces. Here is how to find the right model, use the Inference API, and run models locally with transformers.
Mahmudul Haque Qudrati
CEO & ML Engineer
LangChain is general-purpose and heavy. LlamaIndex is optimized for RAG. Here is when to use each, and when to use neither, with code examples.
Mahmudul Haque Qudrati
CEO & ML Engineer
sentence-transformers, BGE-M3, and Nomic Embed are your main options. Here is how they compare to OpenAI's embeddings and when open source is good enough.
Mahmudul Haque Qudrati
CEO & ML Engineer
SD 1.5, SDXL, SD 3, and Flux are your main model choices. Here is how to run locally, use via API, and integrate image generation in your product.
Mahmudul Haque Qudrati
CEO & ML Engineer
Whisper is the best open source speech-to-text model. Run locally with faster-whisper or free via Groq. Here is how to integrate transcription in Python and Node.js with honest tradeoffs.
Mahmudul Haque Qudrati
CEO & ML Engineer
vLLM, Ollama, and TGI are the main serving options. Here is hardware requirements, latency comparison, and the cost crossover point where self-hosting beats the API.
Mahmudul Haque Qudrati
CEO & ML Engineer
PagedAttention gives vLLM 2-24x throughput over naive implementations. Here is how to set it up, configure batching, quantize models, and calculate hardware costs.
Mahmudul Haque Qudrati
CEO & ML Engineer
QLoRA makes fine-tuning 70B models accessible on a single consumer GPU. Here is the complete setup guide for fine-tuning Llama 3 with Unsloth.
Mahmudul Haque Qudrati
CEO & ML Engineer
MLflow tracks parameters, metrics, and artifacts for ML experiments. Here is how to set it up in 5 lines, integrate with training loops, and when to use it over W&B.
Mahmudul Haque Qudrati
CEO & ML Engineer
DSPy optimizes LLM prompts automatically using your data. Here is when it helps, when it does not, and a complete setup guide for a real use case.
Mahmudul Haque Qudrati
CEO & ML Engineer
CrewAI lets you define agents with roles, assign them tasks, and have them collaborate. Here is when multi-agent beats single-agent, and when it does not.
Mahmudul Haque Qudrati
CEO & ML Engineer