whisper alternatives lately: anyone testing this?

[quiethouse]

What worked for me was pretty simple. If you're on SGLang, I'd confirm the version first and then reduce variables until the behavior is obvious. Once you know the plain path works, then add the faster path back in piece by piece.

[bytegarden73]

I'd approach it in a boring order. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at long context. Not glamorous advice, but repeatable baselines save a lot of time.

[deltagrid]

worth checking the driver first. At this point the real bottleneck is my patience.

[deltastation]

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at speculative decoding.

[amberpatch25]

Interesting, I hadn't tried that. I ended up in about the same place using Ollama. The biggest change for me was simplifying the stack instead of adding one more optimization layer. At this point the real bottleneck is my patience.

[audiopath66]

Appreciate the detail. The pattern around VRAM headroom has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.

[cudascope]

I'd start with the simpler stack.

[deltagrid]

What worked for me was pretty simple. Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at consumer GPUs. Once you know the plain path works, then add the faster path back in piece by piece.

[visionforge]

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

[cudascope]

Interesting, I hadn't tried that. The pattern around NVFP4 has been pretty similar for me. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.