What advantages does this have over taking all your data sources, creating embeddings for all and storing in a vector store. Then a single query would give the top_k responses and a summarization of those will give an answer? I can see that the question explicitly defines the data sources, but struggle to see the utility in this.
Can a subsequent SFT and RTHF with different, additional or lesser contents change the character, improve, or degrade a GPT model? Can you modify a GPT model? How?
Thank you, Jerry and Amog! Nice demo! Keep up the good work!
always great videos! Love this topic but without openai..I would love to see something created only using open source technology...
What advantages does this have over taking all your data sources, creating embeddings for all and storing in a vector store. Then a single query would give the top_k responses and a summarization of those will give an answer? I can see that the question explicitly defines the data sources, but struggle to see the utility in this.
is there a place a beginner can learn about ray from scratch
Can a subsequent SFT and RTHF with different, additional or lesser contents change the character, improve, or degrade a GPT model? Can you modify a GPT model? How?
What’s is the difference ray and embedded chain
Light ide theme 😩 fun video tho
Bro has light theme