10-19-2026, 03:06 AM
That matches what I was seeing. I ended up in about the same place using SGLang. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
trying not to rebuild again
|
comparing models without fooling yourself on GB200
|
|
« Next Oldest | Next Newest »
|
Users browsing this thread: 1 Guest(s)