Mohit Dubey

Gen AI / ML | 14+ years
Youtube - NueralWithMohit:

[Mohit Dubey @MohitMKD]

Say “oh yeah” if you are a human @Uber_Support

[Mohit Dubey @MohitMKD]

Turn your images into knowledge

Create an Image-Aware RAG Pipeline with Cohere Multi-Modal Embeddings & LLaMA 4
https://t.co/yGztXPLu9q

[Mohit Dubey @MohitMKD]

7/
💻 GitHub Code (open-source): 👉 https://t.co/7Ww9ZxioMt
Rerankers = actual relevance
Vector DB = semantic similarity
Both matter. But together? 🔥
Follow me for more GenAI tutorials and insights.
#RAG #VectorDB #Reranker #ChromaDB #LLM #NLP #Python #AI #HuggingFace

[Mohit Dubey @MohitMKD]

Why Rerankers Matter in Vector Search + How to Implement Them

1/
Vector search ≠ relevance.
Just because a document is similar doesn’t mean it’s useful.
That’s why Rerankers matter - and most people skip this step. 🧠
Let me show you why - and how to build one in Python. 👇

[Mohit Dubey @MohitMKD]

6/
📺 Watch the full video tutorial:
👉 https://t.co/wYAfxGBlpa

[Mohit Dubey @MohitMKD]

80% of ML models never make it to production. Why? A model that’s 90% accurate on your machine can drop to <60% with real-world data. Real-world data is noisy & discrete - leading to major performance issues.

#AI #MachineLearning #DataScience #TechChallenges #ML

[Mohit Dubey @MohitMKD]

LoRA has been the most effective & widely used method for fine-tuning LLMs—efficient & production-friendly.

MIT's new paper on Self-Adaptive LLMs shows a shift: models that adapt to new tasks without gradient updates or memory.

📄 https://t.co/duSsB7Fug4
#LLM #AI #LoRA #NLP

[Mohit Dubey @MohitMKD]

7/ 🔄 Don't forget to like, share, and subscribe for more tutorials on AI, ML, and vector databases!

#AI #ChromaDB #VectorSearch #NLP #Embedding #HNSW #MachineLearning #Caching #Python #Tutorial

[Mohit Dubey @MohitMKD]

🧵 Thread: Build a High-Performance Vector Search System with ChromaDB & HNSW!

[Mohit Dubey @MohitMKD]

6/ 🎥 Want to learn the details? Watch the full video here! 👉 Watch the video here! https://t.co/V9DmchHICK

[Mohit Dubey @MohitMKD]

New YouTube tutorial! Learn to build a high-performance vector search system with ChromaDB, HNSW, & MiniLM-L6-v2!

I cover:
✅ Setting up ChromaDB & HNSW
✅ Using MiniLM for text vectorization
✅ Implementing smart query caching

👉https://t.co/V9DmchHICK

#VectorDB #AI #ML