issues: ci - close inactive issue with workflow

[phymbert]

References

• Close inactive issue workflow #6048
• https://docs.github.com/en/actions/managing-issues-and-pull-requests/closing-inactive-issues

[ggerganov]

Let's give it a day to see if there are any concerns about adding this

[ggerganov]

One small concern is that when this runs the first time, all watchers of the repo will receive several hundred notifications about stale issues. Not sure if we can do anything about that though

[ngxson]

Also @ggerganov should we update the pinned issues and pinned discussions? As some pinned discussions are not very up-to-date.

I also think that some PRs deserve more visibility like #5021 , should we have a place to keep track of them (other than "Hot topics" in readme?)

[phymbert]

One small concern is that when this runs the first time, all watchers of the repo will receive several hundred notifications about stale issues. Not sure if we can do anything about that though

Finally it stopped after 30 operations per job, so it will progressivly close stale issues. Great:

• https://github.com/ggerganov/llama.cpp/actions/runs/8311856695/job/22746075694

[ggerganov]

Hm, probably better to close them all at once. Otherwise it will spam us every day for a month

[phymbert]

Just to share that : this last run manages to break/bypass gmail classification ...

Still 500+ to proceed...
Also some people are complaining:

• [Enhancement]: Global probability maximization with backtracking for better results. (aka Beam search) #1392 (comment)
• Ability for ./main to keep the model in memory and pass it more text #23 (comment)

But still I think this is for the good, in a couple of weeks, it will become BAU.

[slaren]

Funnily enough, these issues are truly stale: beam search has been implemented for ages and keeping the model in memory to pass it more text is what the server does. It's useless to have thousands of open issues, it just makes developers ignore the list of issues.