Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Free Dolly: First truly open instruction-tuned LLM

databricks.com

Post reply on HN
Print view View original

61–70 of 70 posts

Re: Free Dolly: First truly open instruction-tuned LLM

#61

Post by rxin »

alsodumb wrote: post #43
nidnogg wrote: post #28

There's some blatant astroturfing from new accounts going on in this thread - I gotta say it's not the best impression

Maybe OP, Data bricks founder, should reply to this comment and answer if he or anyone else in his team paid people to get this post with a not-so-useful-dataset to the top of HN.

(OP / Databricks cofounder here)

I'm not sure what happened but I just sent an email out internally to ask people not to do this. The team might have gotten overly excited by this because they were all part of the creation of the dataset and the model.

Re: Free Dolly: First truly open instruction-tuned LLM

#62

Post by mydpy »

covi wrote: post #59

Kudos to Databricks! Anyone has insights into benchmark & real-world quality? From https://huggingface.co/databricks/dolly-v2-12b#benchmark-met... , it seems like dolly-v2-12b's benchmark results are actually slightly worse than dolly-v1-6b. A commercially viable instruction-tuned LLM is still a huge deal.

Right, but this is still impressive given how quickly Databricks created and open-sourced the dataset. I would expect more improvements in the future.

Re: Free Dolly: First truly open instruction-tuned LLM

#63

Post by visarga »

IanCal wrote: post #55
visarga wrote: post #23

Earlier quoted context omitted.

My hunch is that OpenAI's window of supremacy will be short. Even if they keep being SOTA, the open sourced models will eat away market underneath them. By the next year they will only be able to sell GPT-4 or 5. At some point the open models will be good enough for 99% of use cases.

Quality is one aspect, running them is another. If I've already got everything setup with them and they work efficiently, they could also offer open source models and let me pay for usage. Both bursty usage and low constant usage benefit from paying per token and having some shared & large infrastructure to use. I don't want to be running a bunch of h100s, I just want my requests processed. If they're selling gpt-5 a…

> I don't want to be running a bunch of h100s, I just want my requests processed.

Assuming you don't have sensitive data and that you never try anything outside the rules.

Re: Free Dolly: First truly open instruction-tuned LLM

#65

Post by IanCal »

visarga wrote: post #63
IanCal wrote: post #55

Earlier quoted context omitted.

Quality is one aspect, running them is another. If I've already got everything setup with them and they work efficiently, they could also offer open source models and let me pay for usage. Both bursty usage and low constant usage benefit from paying per token and having some shared & large infrastructure to use. I don't want to be running a bunch of h100s, I just want my requests processed. If they're selling gpt-5 a…

> I don't want to be running a bunch of h100s, I just want my requests processed. Assuming you don't have sensitive data and that you never try anything outside the rules.

Those are hosting issues really and shouldn't be an issue for most companies.

I'd be absolutely shocked if they don't launch a version where it's run on more secure setups, particularly as they've got huge Microsoft backing.

Re: Free Dolly: First truly open instruction-tuned LLM

#66

Post by Nifty3929 »

ibejoeb wrote: post #45

I deal with this stuff daily, so I think it's probably irrationally grinding my gears, but: >> How do I build a campfire? > Safety should always come first when starting a campfire. Hold up: should I touch the fire? It doesn't say. OK, there's perfectly legitimate advice in the output, like "have water nearby," but give me a break already. They're finetuning for commercial application. If I'm building business tools,…

This is just being prudent from a business perspective. Much to lose due to lawsuits. Look at the warnings on Q-Tips (which everybody in the US uses to dry their ears) which have warnings not to use in ears.

Re: Free Dolly: First truly open instruction-tuned LLM

#67

Post by baidifnaoxi »

rxin wrote: post #61
alsodumb wrote: post #43

Earlier quoted context omitted.

Maybe OP, Data bricks founder, should reply to this comment and answer if he or anyone else in his team paid people to get this post with a not-so-useful-dataset to the top of HN.

(OP / Databricks cofounder here) I'm not sure what happened but I just sent an email out internally to ask people not to do this. The team might have gotten overly excited by this because they were all part of the creation of the dataset and the model.

This seems to happen a bit too often tbh on all things Databricks.

Re: Free Dolly: First truly open instruction-tuned LLM

#68

Post by lostmsu »

pauldix wrote: post #54

Shame that this is flagged, I think this is a really exciting development and was hoping to see the discussion around it. Open sourcing the fine tuning training set is a great building block. Will be exciting to see if others continue to build on this. More open source datasets, models, and evaluation frameworks will accelerate the development and adoption of LLMs. It adds more hackers to the mix building the core, r…

There are already a number of instruction-tuning datasets.

This announcement does not provide any benchmarks, so it is impossible to tell how useful the model is.

Re: Free Dolly: First truly open instruction-tuned LLM

#70

Post by IanCal »

IanCal wrote: post #65
visarga wrote: post #63

Earlier quoted context omitted.

> I don't want to be running a bunch of h100s, I just want my requests processed. Assuming you don't have sensitive data and that you never try anything outside the rules.

Those are hosting issues really and shouldn't be an issue for most companies. I'd be absolutely shocked if they don't launch a version where it's run on more secure setups, particularly as they've got huge Microsoft backing.

You can run this in azure which I feel solves basically all of the hosting issues.
Post reply on HN