This about sums up my experiences with LLM generated code. One thing I’ve seen a lot is during the debugging stage, it supposedly finds the problem, provides new code, new code has same/new problem, paste error, generates new code that has the original problem. Basically it goes into debugging loops. Sometimes it proposes a new method that changes code across multiple files and has a host of other issues. I think at some point it’s context window is full and loses the plot. It’s like working with someone who may have been an expert at programming but got hit by a bus and has brain damage.

I created a free AI Implementation Handbook to help teams immediately identify where to start implementing AI in their processes: www.makethejump.ai/ai-implementation-handbook

I created a free AI Implementation Handbook to help teams immediately identify where to start implementing AI in their processes: www.makethejump.ai/ai-implementation-handbook

Congratulations Sam 🎉it’s been so good learn from you ❤

Some great points! It’s encouraging to see content on this subject.

amazing video!

This video was nice, can you please explain how to create custom model for language translation using transformer in hugging face.

This video was nice, can you please explain how to create custom model for language translation using transformer in hugging face.

I subscribed because of the prayer. You're not ashamed of your faith in Jesus. Keep it up

讲得很清楚，informative

thank you so much

You really simplified it. Thank you. subscribed.

You just gained a subscriber! Thanks!

Thank you sir for provding comprehensive video on creating dataset , i was tired of watching many video but i was not getting the right one , but the level of explaination you have provided is awesome . looking many further video in future from end