hybrid retrieval results

[tensorfield88]

Spent some time on this today and wanted to compare notes.

I've gone back and forth between PyTorch and TensorRT-LLM for Kimi. The interesting difference for me has been workstation thermals.

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

[sanepeak48]

A couple questions before I guess too hard. Do you see the same thing with a smaller batch or shorter prompt?

[tinycontext]

following this because I'm in the same boat.

[homelabhenry]

That matches what I was seeing. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one more optimization layer.

[bytelab]

That matches what I was seeing. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding one mo

Good point. I ended up in about the same place using MLX. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

[tensorfield88]

[quote]That matches what I was seeing. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of adding

Interesting, I hadn't tried that. I don't think your conclusion is far off.

[batchshed]

[quote][quote]That matches what I was seeing. I don't think your conclusion is far off. The biggest change for me was simplifying the stack instead of

What worked for me was pretty simple. On 4090, the easiest win is often to back off the aggressive settings and see whether the problem survives.

[tinycontext]

I'd approach it in a boring order. On Threadripper box, the easiest win is often to back off the aggressive settings and see whether the problem survives. Once you know the plain path works, then add the faster path back in piece by piece.

If you post your exact settings, people here can probably narrow it down pretty fast.

[homelabhenry]

Thanks, that helps. That matches what I saw on B200. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing. Turns out 'just one more test' was a lie.

[batchshed]

A couple questions before I guess too hard. What model, what runtime, and roughly what context size?