Thanx for your kind comment. I'll try to bring a lot more like this ... plz like, subscribe and tell others. It's not easy to keep doing this with a full-time job! Take care...
Great work brother, just wanted to ask how can i implement chaining in this and what library should i use to make a good looking UI similar to a chatbox
Thanx for your kind comment. If you're looking to implement chaining I'd strongly recommend looking into my Langchain series. There I talk about LLM Chain, Sequential Chain etc. And the great thing about Langchain is that you can use any model with it ... you don't have to be stuck to OpenAI only ... even though that's the best so far ... esp, gpt-4 ... also, as far as good looking UI goes, you can consider Streamlit. For a python dev, you don't even need to know html, css etc. Soon I'll be starting a series on that ... as soon as I can find some time ... you know ... I gotta full-time job ... but plz subscribe and keep an eye on this channel. I'd like bring in everything I learn about AI in here ... you will know everything I know bro ... that's the goal. Thanx again.
@@htmlfivedev appreciate the advice, I'm also starting my career as an AI engineer, I'm the only AI engineer in my company and they want me to build an LLM model which is fine tuned based on their company's data if the prompt doesn't find anything in company's data then it should move to any LLM model, I've only had experience of few months I'm stuck in this situation trying to understand all these advance topics in a short period of time, your video really helped and thanks for replying hope you achieve the success you really deserve, if you have any advice how I should I bee doing the task I mentioned above, a little roadmap would really help a lot. Thanks
Ok, for what you are trying to do I would recommend (strongly) to go the RAG bot route rather than fine tuning ... I have tried FT but it doesn't work! Plus, it is more expensive. I suggest you what my videos on RAG and find out more on the topic. This way you will have a prototype in no time.
