Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
hybrid retrieval results for FLUX
#1
Posting this in case it saves someone else a little time.

I'm not trying to build a perfect architecture diagram. I just want a sane starting point around Ollama and GB200.

I'm especially interested in mistakes worth avoiding.

This is mostly for local work rather than public serving.
rack in the garage • local models everywhere
Reply
#2
A couple questions before I guess too hard. Is this local-only or are you routing anything externally?
Reply
#3
I hit almost the same issue last night.
rack in the garage • local models everywhere
Reply
#4
Can you fill in a bit more context? What model, what runtime, and roughly what context size?
Reply
#5
subscribed, curious where this lands. I keep promising myself this is the last rebuild.
Reply
#6
I'd approach it in a boring order. Before rebuilding anything, I'd test one clean baseline with Kimi and compare from there. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
rack in the garage • local models everywhere
Reply
#7
Can you fill in a bit more context? Do you see the same thing with a smaller batch or shorter prompt?
Reply
#8
that fixed it for me too.

I'm happy to test one more variant if you want a second data point.
Reply
#9
What exact setup are you running? Do you see the same thing with a smaller batch or shorter prompt?
Reply
#10
worth checking the driver first.

If you post your exact settings, people here can probably narrow it down pretty fast.
context windows are not a product strategy
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)