From the course: LLMOps in Practice: A Deep Dive
Unlock the full course today
Join today to access over 24,900 courses taught by industry experts.
Demonstrating the ops project completed - Python Tutorial
From the course: LLMOps in Practice: A Deep Dive
Demonstrating the ops project completed
Okay. So here's the chatbot that we've been working on and we've seen logging already. But there's a couple of extra features I'd like you to see. So if I said, for example, if I'm interested in the topic of the migration patterns of wombats, we'll see the usual, the chatbot does its thing. It's thinking it's sending that to the back end OpenAI. And it should give me back a response. But we'll see something new this time. So now we can see these boxes: good, neutral, or bad at the bottom of the screen. With the idea being that if it's good, I'll continue. If it's neutral, I'll still continue, but I'll just log that the user, hey, wasn't very overwhelmed by the topic, but if it's bad, what will happen is I don't like that response, so it will trigger another one. So it goes into the thinking cycle and it will give me back a different answer because I didn't really like this one. So again, we're beginning to get the details here for feedback from a human so that we can have that…
Contents
-
-
-
-
(Locked)
Coding for logging7m 35s
-
(Locked)
Exploring the logging system4m 23s
-
(Locked)
RLHF and user feedback1m 52s
-
(Locked)
Challenge: Implementing RLHF and user feedback2m 35s
-
(Locked)
Demonstrating the ops project completed2m 40s
-
(Locked)
Solution: Completing an ops project5m 2s
-
(Locked)
Demonstrating the code for the ops17m 12s
-
(Locked)
-
-
-