You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There's nothing that I hate more than a bunch of results without any proof. So either you researchers are too lazy to credibly back up what it is you claimed to have designed or you all simply got the instruction dataset directly from OpenAI or some other entity.
Don't parade around as though you're contributing to open source when you're really part of the circle jerk of corporate shills that run around touting 'open source' while not adhering to any of the standards or principles of open source. Repos like this are poison in the community because they send the message to other researchers that its okay to move like this.
And it isn't.
For those awaiting a viable repo that breaks down how to generate the synthetic data you're looking, I got that on the way. No half assing or pretending like I do. I have an actual repo that's on the way for you all that will thoroughly explain - line by line - exactly how this synthetic data is to be created and how you are supposed to augment it with the evol-instruct method so that we can put the power in the hands of more people rather than keeping all of the secrets behind a lock and key.
I'm not afraid to tell you my methodologies because it is my hope that others will be able to replicate my process and improve upon what I've already iterated. If that's not happening then we all lose because we fail to have an educated, balanced discussion about the true losses that the open source community takes when assholes like the ones that launched this repo pretend like they're here to actually help and not harm.
The text was updated successfully, but these errors were encountered:
There's nothing that I hate more than a bunch of results without any proof. So either you researchers are too lazy to credibly back up what it is you claimed to have designed or you all simply got the instruction dataset directly from OpenAI or some other entity.
Don't parade around as though you're contributing to open source when you're really part of the circle jerk of corporate shills that run around touting 'open source' while not adhering to any of the standards or principles of open source. Repos like this are poison in the community because they send the message to other researchers that its okay to move like this.
And it isn't.
For those awaiting a viable repo that breaks down how to generate the synthetic data you're looking, I got that on the way. No half assing or pretending like I do. I have an actual repo that's on the way for you all that will thoroughly explain - line by line - exactly how this synthetic data is to be created and how you are supposed to augment it with the
evol-instruct
method so that we can put the power in the hands of more people rather than keeping all of the secrets behind a lock and key.I'm not afraid to tell you my methodologies because it is my hope that others will be able to replicate my process and improve upon what I've already iterated. If that's not happening then we all lose because we fail to have an educated, balanced discussion about the true losses that the open source community takes when assholes like the ones that launched this repo pretend like they're here to actually help and not harm.
The text was updated successfully, but these errors were encountered: