to a specific task by using a small dataset
also leaarned about lora and qlora
ye sab thik hai
kal anthropic ne ek aur model daal diya ghante ka "we must pace the frontier"
but anyways its cool will try to build something
Day 10. Just posting here what I'm learning: the journey of LLMs so far has started from preโtraining, prompting, RAG, agents, and fineโtuning.
Instead of spending millions of dollars to pre train a model on trillions of tokens from scratch, fine tuning takes a pre trained model