https://www.datasculpting.ai/inference/ Mon, 26 Oct 2026 11:43:48 +0000 MyBB https://www.datasculpting.ai/inference/showthread.php?tid=1490 Sun, 18 Oct 2026 07:23:40 +0000 ggufvault]]> https://www.datasculpting.ai/inference/showthread.php?tid=1490
The more I work with local setups, the more it seems like long context decides everything downstream.

Curious whether that lines up with what other people are seeing.]]>
The more I work with local setups, the more it seems like long context decides everything downstream.

Curious whether that lines up with what other people are seeing.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1310 Thu, 15 Oct 2026 16:23:30 +0000 dockworks23]]> https://www.datasculpting.ai/inference/showthread.php?tid=1310
One thing I keep noticing is that people underestimate tool calling until they actually try to run this stuff for more than one afternoon.

Maybe obvious to some of you, but it clicked for me this week. At this point the real bottleneck is my patience.]]>
One thing I keep noticing is that people underestimate tool calling until they actually try to run this stuff for more than one afternoon.

Maybe obvious to some of you, but it clicked for me this week. At this point the real bottleneck is my patience.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1225 Thu, 15 Oct 2026 02:56:11 +0000 northgrid33]]> https://www.datasculpting.ai/inference/showthread.php?tid=1225
Short version: vLLM behaved well for me, but only after I stopped ignoring VRAM headroom.

Curious if anyone else landed in the same place.]]>
Short version: vLLM behaved well for me, but only after I stopped ignoring VRAM headroom.

Curious if anyone else landed in the same place.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1082 Sun, 11 Oct 2026 22:36:31 +0000 oxidegrove40]]> https://www.datasculpting.ai/inference/showthread.php?tid=1082
My current setup is Ollama with 4090, and I keep running into tool calling.

Not looking for the perfect answer, mostly what worked in real use.]]>
My current setup is Ollama with 4090, and I keep running into tool calling.

Not looking for the perfect answer, mostly what worked in real use.]]> https://www.datasculpting.ai/inference/showthread.php?tid=768 Mon, 05 Oct 2026 22:34:20 +0000 orbitgarden]]> https://www.datasculpting.ai/inference/showthread.php?tid=768
Short version: MLX behaved well for me, but only after I stopped ignoring prefill/decode split.

Curious if anyone else landed in the same place.]]>
Short version: MLX behaved well for me, but only after I stopped ignoring prefill/decode split.

Curious if anyone else landed in the same place.]]> https://www.datasculpting.ai/inference/showthread.php?tid=744 Sun, 04 Oct 2026 20:30:07 +0000 llm_wren]]> https://www.datasculpting.ai/inference/showthread.php?tid=744
I've gone back and forth between Ollama and Ollama for Kimi. The interesting difference for me has been router policy.

What would you pick if you had to live with it for a few months?

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]>
I've gone back and forth between Ollama and Ollama for Kimi. The interesting difference for me has been router policy.

What would you pick if you had to live with it for a few months?

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]> https://www.datasculpting.ai/inference/showthread.php?tid=712 Sun, 04 Oct 2026 01:55:22 +0000 thetanest]]> https://www.datasculpting.ai/inference/showthread.php?tid=712
I tried this with Kimi using vLLM on GB200. The headline result was better than I expected, but VRAM headroom still ended up being the limiting factor.

Happy to post more details if anyone wants them.]]>
I tried this with Kimi using vLLM on GB200. The headline result was better than I expected, but VRAM headroom still ended up being the limiting factor.

Happy to post more details if anyone wants them.]]> https://www.datasculpting.ai/inference/showthread.php?tid=688 Fri, 02 Oct 2026 22:15:10 +0000 llm_wren]]> https://www.datasculpting.ai/inference/showthread.php?tid=688
I'm not trying to build a perfect architecture diagram. I just want a sane starting point around llama.cpp and 4090.

I'm especially interested in mistakes worth avoiding.

This is mostly for local work rather than public serving.]]>
I'm not trying to build a perfect architecture diagram. I just want a sane starting point around llama.cpp and 4090.

I'm especially interested in mistakes worth avoiding.

This is mostly for local work rather than public serving.]]> https://www.datasculpting.ai/inference/showthread.php?tid=625 Thu, 01 Oct 2026 05:21:00 +0000 cudaline]]> https://www.datasculpting.ai/inference/showthread.php?tid=625
This isn't revolutionary, but the combination of TensorRT-LLM and dual 3090 felt surprisingly usable once I tuned around embeddings.

Feel free to tear it apart if you see something dumb.]]>
This isn't revolutionary, but the combination of TensorRT-LLM and dual 3090 felt surprisingly usable once I tuned around embeddings.

Feel free to tear it apart if you see something dumb.]]> https://www.datasculpting.ai/inference/showthread.php?tid=507 Sat, 26 Sep 2026 15:44:51 +0000 amberplane]]> https://www.datasculpting.ai/inference/showthread.php?tid=507
I'm mostly working with Qwen in llama.cpp on 5090. The part I'm trying to understand is NVFP4.

Would appreciate practical advice rather than theory.]]>
I'm mostly working with Qwen in llama.cpp on 5090. The part I'm trying to understand is NVFP4.

Would appreciate practical advice rather than theory.]]> https://www.datasculpting.ai/inference/showthread.php?tid=496 Fri, 25 Sep 2026 21:25:09 +0000 ggufvault]]> https://www.datasculpting.ai/inference/showthread.php?tid=496
I'm mostly working with Llama in LM Studio on A6000. The part I'm trying to understand is tail latency.

Not looking for the perfect answer, mostly what worked in real use.]]>
I'm mostly working with Llama in LM Studio on A6000. The part I'm trying to understand is tail latency.

Not looking for the perfect answer, mostly what worked in real use.]]> https://www.datasculpting.ai/inference/showthread.php?tid=470 Fri, 25 Sep 2026 04:29:44 +0000 oxidegrove40]]> https://www.datasculpting.ai/inference/showthread.php?tid=470
I tried this with Llama using vLLM on GB200. The headline result was better than I expected, but KV cache pressure still ended up being the limiting factor.

Curious if anyone else landed in the same place.

This is mostly for local work rather than public serving.]]>
I tried this with Llama using vLLM on GB200. The headline result was better than I expected, but KV cache pressure still ended up being the limiting factor.

Curious if anyone else landed in the same place.

This is mostly for local work rather than public serving.]]> https://www.datasculpting.ai/inference/showthread.php?tid=463 Fri, 25 Sep 2026 02:01:18 +0000 llm_wren]]> https://www.datasculpting.ai/inference/showthread.php?tid=463
Maybe useful as a starting point: MLX on 4090, keeping NVFP4 in mind from the start.

Feel free to tear it apart if you see something dumb.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]>
Maybe useful as a starting point: MLX on 4090, keeping NVFP4 in mind from the start.

Feel free to tear it apart if you see something dumb.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]> https://www.datasculpting.ai/inference/showthread.php?tid=465 Fri, 25 Sep 2026 00:52:00 +0000 bytegarden73]]> https://www.datasculpting.ai/inference/showthread.php?tid=465
Two paths both seem viable, but they break down differently once tail latency becomes important.

Interested in tradeoffs more than fan loyalty.

This is mostly for local work rather than public serving.]]>
Two paths both seem viable, but they break down differently once tail latency becomes important.

Interested in tradeoffs more than fan loyalty.

This is mostly for local work rather than public serving.]]> https://www.datasculpting.ai/inference/showthread.php?tid=275 Tue, 15 Sep 2026 01:00:43 +0000 llm_wren]]> https://www.datasculpting.ai/inference/showthread.php?tid=275
I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling HBM constraints.

If you've tested something similar, what did you keep and what did you stop bothering with?]]>
I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling HBM constraints.

If you've tested something similar, what did you keep and what did you stop bothering with?]]>