10-02-2026, 05:02 PM
Did a small comparison this afternoon. On dual 3090 with SGLang, I saw basically the same behavior. The surprise was that memory pressure showed up before throughput collapsed.
small models can still be good
|
benchmark result I don't trust with LM Studio
|
|
« Next Oldest | Next Newest »
|
| Messages In This Thread |
|
benchmark result I don't trust with LM Studio - by promptspark58 - 10-02-2026, 03:51 PM
Re: benchmark result I don't trust with LM Studio - by localtinker - 10-02-2026, 05:02 PM
|
Users browsing this thread: 1 Guest(s)