Evidence receipt / evaluation
Published · transcript-backedAlistair Pullen: evaluation
22 Aug 2024 Latent Space Is finetuning GPT4o worth it? — with Alistair Pullen, Cosine (Genie)
“It's hilarious, like we had to filter out so much of that stuff because when we first did the 16k model, like the amount of readme updating that went in, we did like no data cleaning, no real, like, we just sort of threw it in and saw what happened. And it was just like, It was really good at updating readme, it was really good at writing some comments, really good at, um, complaining in Git reviews, in PR reviews, rather, and it would, again, like, we didn't clean the data, so you'd, like, give it some feedback, and it would just, like, reply, and, like, it would just be quite insubordinate when it was getting back to you, like, no, I don't think you're right, and it would just sort of argue with you, so The process of doing all that was super interesting because we realized from the beginning, okay, there's a huge amount of work that needs to go into like cleaning this, getting it aligned with what we want the model to do to be able to get the model to be useful in some way.”
Source trail
Everything needed to verify it.
- Speaker
- Alistair Pullen
- Attribution
- Verified speaker
- Claim type
- evaluation
- Recorded
- 22 Aug 2024
- Publisher
- Latent Space
Transcript context
…How does that differ when you work with customers that have private workflows? Like, do you think, is there usually a big delta between what you get in open source and maybe public data versus like Yeah, yeah, yeah. When you scrape enough of it, most of open source is updating readmes and docs. It's hilarious, like we had to filter out so much of that stuff because when we first did the 16k model, like the amount of readme updating that went in, we did like no data cleaning, no real, like, we just sort of threw it in and saw what happened. And it was just like, It was really good at updating readme, it was really good at writing some comments, really good at, um, complaining in Git reviews, in PR reviews, rather, and it would, again, like, we didn't clean the data, so you'd, like, give it some feedback, and it would just, like, reply, and, like, it would just be quite insubordinate when it was getting back to you, like, no, I don't think you're right, and it would just sort of argue with you, so The process of doing all that was super interesting because we realized from the beginning, okay, there's a huge amount of work that needs to go into like cleaning this, getting it aligned with what we want the model to do to be able to get the model to be useful in some way. I'm curious, like, how do you think about the customer willingness? To share all of this historical data, I've done a lot of developer tools investing in my career and getting access to the code base is always one of the hard things. Are people getting more cautious about sharing this information? In the past, it was maybe like, you know, you're using static analysis tool, like whatever else you need to plug into the code base, fine. Now you're building. A model based on it, like, uh, what's the discussion going into these companies? Are most people comfortable with, like, letting you see how to work and sharing everything?…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.