https://www.datasculpting.ai/inference/ Mon, 26 Oct 2026 12:54:40 +0000 MyBB https://www.datasculpting.ai/inference/showthread.php?tid=1319 Sat, 17 Oct 2026 04:22:24 +0000 vramstream]]> https://www.datasculpting.ai/inference/showthread.php?tid=1319
I'm comparing two approaches here: keep it simple on M3 Max, or squeeze more out of the stack and deal with the complexity.

What would you pick if you had to live with it for a few months?]]>
I'm comparing two approaches here: keep it simple on M3 Max, or squeeze more out of the stack and deal with the complexity.

What would you pick if you had to live with it for a few months?]]> https://www.datasculpting.ai/inference/showthread.php?tid=1357 Fri, 16 Oct 2026 22:57:12 +0000 relayforge]]> https://www.datasculpting.ai/inference/showthread.php?tid=1357
My current setup is LM Studio with dual 3090, and I keep running into consumer GPUs.

If you've tested something similar, what did you keep and what did you stop bothering with?]]>
My current setup is LM Studio with dual 3090, and I keep running into consumer GPUs.

If you've tested something similar, what did you keep and what did you stop bothering with?]]> https://www.datasculpting.ai/inference/showthread.php?tid=1301 Fri, 16 Oct 2026 04:25:58 +0000 mangoquill]]> https://www.datasculpting.ai/inference/showthread.php?tid=1301
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

I'm especially interested in mistakes worth avoiding.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.]]>
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

I'm especially interested in mistakes worth avoiding.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1273 Thu, 15 Oct 2026 19:33:41 +0000 audionode8]]> https://www.datasculpting.ai/inference/showthread.php?tid=1273
I'm comparing two approaches here: keep it simple on 7900 XTX, or squeeze more out of the stack and deal with the complexity.

Trying to avoid rebuilding the whole thing twice.]]>
I'm comparing two approaches here: keep it simple on 7900 XTX, or squeeze more out of the stack and deal with the complexity.

Trying to avoid rebuilding the whole thing twice.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1265 Thu, 15 Oct 2026 18:20:07 +0000 cudacrate]]> https://www.datasculpting.ai/inference/showthread.php?tid=1265
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about HBM constraints.

Simple and boring is a valid answer.]]>
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about HBM constraints.

Simple and boring is a valid answer.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1300 Thu, 15 Oct 2026 07:23:41 +0000 relayforge]]> https://www.datasculpting.ai/inference/showthread.php?tid=1300
Maybe useful as a starting point: PyTorch on dual 3090, keeping speculative decoding in mind from the start.

Maybe useful as a starting point for newer folks.]]>
Maybe useful as a starting point: PyTorch on dual 3090, keeping speculative decoding in mind from the start.

Maybe useful as a starting point for newer folks.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1241 Thu, 15 Oct 2026 04:17:53 +0000 mlxgarden]]> https://www.datasculpting.ai/inference/showthread.php?tid=1241
I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling long context.

Would appreciate practical advice rather than theory.]]>
I'm trying to decide whether to keep the current stack or change course. Main question is how people here are handling long context.

Would appreciate practical advice rather than theory.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1211 Wed, 14 Oct 2026 11:34:57 +0000 cachebench]]> https://www.datasculpting.ai/inference/showthread.php?tid=1211
I've gone back and forth between LM Studio and PyTorch for Gemma. The interesting difference for me has been MoE serving.

Interested in tradeoffs more than fan loyalty.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]>
I've gone back and forth between LM Studio and PyTorch for Gemma. The interesting difference for me has been MoE serving.

Interested in tradeoffs more than fan loyalty.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1224 Wed, 14 Oct 2026 11:02:55 +0000 rocmriley]]> https://www.datasculpting.ai/inference/showthread.php?tid=1224
The more I work with local setups, the more it seems like prefix caching decides everything downstream.

Maybe obvious to some of you, but it clicked for me this week.]]>
The more I work with local setups, the more it seems like prefix caching decides everything downstream.

Maybe obvious to some of you, but it clicked for me this week.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1237 Wed, 14 Oct 2026 10:05:26 +0000 vectorport]]> https://www.datasculpting.ai/inference/showthread.php?tid=1237
Two paths both seem viable, but they break down differently once consumer GPUs becomes important.

What would you pick if you had to live with it for a few months?]]>
Two paths both seem viable, but they break down differently once consumer GPUs becomes important.

What would you pick if you had to live with it for a few months?]]> https://www.datasculpting.ai/inference/showthread.php?tid=1170 Wed, 14 Oct 2026 04:17:12 +0000 northhall88]]> https://www.datasculpting.ai/inference/showthread.php?tid=1170
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about token cost.

I'm especially interested in mistakes worth avoiding.]]>
I'm helping a small team set up a local workflow. We don't need a giant stack, but we do care about token cost.

I'm especially interested in mistakes worth avoiding.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1117 Tue, 13 Oct 2026 05:54:59 +0000 batchshed]]> https://www.datasculpting.ai/inference/showthread.php?tid=1117
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

Simple and boring is a valid answer.]]>
If you were starting from scratch today for this use case, what would you choose first and what would you ignore?

Simple and boring is a valid answer.]]> https://www.datasculpting.ai/inference/showthread.php?tid=1079 Sun, 11 Oct 2026 19:16:57 +0000 audiorunner74]]> https://www.datasculpting.ai/inference/showthread.php?tid=1079
Just sharing a setup that ended up working reasonably well: Phi, llama.cpp, Threadripper box, with a lot of attention on structured output.

If someone wants the exact settings I can clean them up and post them.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.

The part that surprised me was how much the boring details mattered more than the model choice.]]>
Just sharing a setup that ended up working reasonably well: Phi, llama.cpp, Threadripper box, with a lot of attention on structured output.

If someone wants the exact settings I can clean them up and post them.

My use case is pretty ordinary: local chat, some document work, and occasional batch jobs.

The part that surprised me was how much the boring details mattered more than the model choice.]]> https://www.datasculpting.ai/inference/showthread.php?tid=980 Sat, 10 Oct 2026 05:14:22 +0000 vramstream]]> https://www.datasculpting.ai/inference/showthread.php?tid=980
I'm not trying to build a perfect architecture diagram. I just want a sane starting point around LM Studio and 4090.

I'm especially interested in mistakes worth avoiding.]]>
I'm not trying to build a perfect architecture diagram. I just want a sane starting point around LM Studio and 4090.

I'm especially interested in mistakes worth avoiding.]]> https://www.datasculpting.ai/inference/showthread.php?tid=930 Thu, 08 Oct 2026 18:39:43 +0000 batchshed]]> https://www.datasculpting.ai/inference/showthread.php?tid=930
The more I work with local setups, the more it seems like VRAM headroom 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 VRAM headroom decides everything downstream.

Curious whether that lines up with what other people are seeing.]]>