Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
multimodal local stack that feels real on GB200
#1
Did a bit of testing and thought I'd post the result here.

Short version: MLX behaved well for me, but only after I stopped ignoring reranking.

Happy to post more details if anyone wants them. At this point the real bottleneck is my patience.
Reply
#2
Quote:Did a bit of testing and thought I'd post the result here.

Tried to reproduce it just now. The issue went away once I changed one setting and restarted the stack cleanly. The main thing that moved was latency, not quality.
Reply
#3
Did a small comparison this afternoon. On M3 Max with LM Studio, I saw basically the same behavior. The main thing that moved was latency, not quality.
Reply
#4
Quote:Did a small comparison this afternoon. On M3 Max with LM Studio, I saw basically the same behavior. The main thing that moved was latency, not quality

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at small local stacks. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
Reply
#5
I'd approach it in a boring order. On A6000, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it. I keep promising myself this is the last rebuild.
image models by night, local pipelines by day
Reply
#6
A couple questions before I guess too hard. Is this local-only or are you routing anything externally?
Reply
#7
I tested this tonight. The issue went away once I changed one setting and restarted the stack cleanly.
Reply
#8
Quote:I tested this tonight. The issue went away once I changed one setting and restarted the stack cleanly.

A couple things I'd check first. On 4090, the easiest win is often to back off the aggressive settings and see whether the problem survives. A lot of these threads end up being one bad config plus a pile of conclusions drawn from it.
Reply
#9
I hit almost the same issue last night.
small models can still be good
Reply
#10
What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at tail latency.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)