InTowards AIbysaniya jaswani·Jun 16If Your Model Inference is Slow, MOE Can Fix it“Mixture of Experts makes model inference faster. To scale request volume, MoE optimizes token routing.”
InTowards AIbysaniya jaswani·May 23Which RAG Works for You in Production?A guide to naive RAG, advanced retrieval strategies, Flare-RAG, GraphRAG, and agentic pipelines, and how to create your architecture.A response icon1A response icon1
InTowards AIbysaniya jaswani·May 214 Memory Types that Your Agent Must HaveIf you do not design your agent like this, it will fail in production.A response icon1A response icon1
saniya jaswani·Mar 5LLM Inference: From Theory to Cost Effective DeploymentLLM inference is not normal deep learning model deployment nor is it trivial when it comes to managing scale, performance and COST…
saniya jaswani·Feb 14How to scale your LLM training — Last partTaking 70B Model, what GPU would i choose and what would be my batch size.
saniya jaswani·Feb 13How to scale your LLM training — Part 4Cool stuff about, what would technique you would be doing for training.
saniya jaswani·Feb 12How to scale your LLM training — Part 3Estimate exactly how much time would take to complete training given Parameter and Database size
saniya jaswani·Feb 11How to scale your LLM training — Part 2Structure of TPU & how does data move in TPU
saniya jaswani·Feb 10How to scale your LLM trainingHow to choose right GPU and number of GPUs , and scale your LLM with more data.