+- Inference (https://www.datasculpting.ai/inference)
+-- Forum: Harnesses, Agents & Orchestration (https://www.datasculpting.ai/inference/forumdisplay.php?fid=11)
+--- Forum: Workflows / Automation (https://www.datasculpting.ai/inference/forumdisplay.php?fid=47)
+--- Thread: DeepSeek V4.1 impressions for Llama (/showthread.php?tid=1464)
Pages:
1
2
Re: DeepSeek V4.1 impressions for Llama - rocmhouse90 - 10-18-2026
Quote:Quote:that fixed it for me too.
Thanks, that helps. I ended up in about the same place using PyTorch. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.
Re: DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
I don't think that's the whole story. I get why that seems right, but on dual 3090 the bottleneck felt elsewhere. Not saying you're wrong, just that I wouldn't lock that in as the lesson yet.
Re: DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
following this because I'm in the same boat.
Re: DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
that fixed it for me too.
Re: DeepSeek V4.1 impressions for Llama - mangofield33 - 10-18-2026
Quote:following this because I'm in the same boat.
Interesting, I hadn't tried that. That matches what I saw on A6000. The biggest change for me was simplifying the stack instead of adding one more optimization layer.
Re: DeepSeek V4.1 impressions for Llama - cozysmith - 10-18-2026
Quote:I don't think that's the whole story. I get why that seems right, but on dual 3090 the bottleneck felt elsewhere. Not saying you're wrong, just that I
I don't think that's the whole story. I get why that seems right, but on 5090 the bottleneck felt elsewhere.
Re: DeepSeek V4.1 impressions for Llama - bytesupply76 - 10-18-2026
Ran a quick check on my box. My result was mixed: first run looked bad, second run looked normal. I expected the model choice to matter more than it did.
Re: DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
A couple things I'd check first. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at prefix caching. Once you know the plain path works, then add the faster path back in piece by piece.
DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
Posting this in case it saves someone else a little time.
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about small local stacks.
Simple and boring is a valid answer.
My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.
Re: DeepSeek V4.1 impressions for Llama - routecraft - 10-18-2026
Quote:that fixed it for me too.
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.