Gabriel Grant

ideally would be good to figure out why it's blowing up with the latest vllm rather than pegging to an old version. do you have any insights?

saw this PR only removed the old HF chat link. would it make sense to add a link to this colab into the readme?

AFAIU ORT has merged WebGPU support: https://github.com/microsoft/onnxruntime/issues/11695 What's needed to take advantage of this on the transformers.js side?

Unfortunately the WebGPU implementation is currently slower than the WASM version, though: https://github.com/microsoft/onnxruntime/issues/18754#issuecomment-1859722309 Would be great to know what's needed to support WebGPU in transformers.js assuming that perf issue gets...

@DavidGOrtega that's great news! to be clear, are you running your models directly on ORT? or using JSEP through transformers.js somehow? would love to hear more details about exactly what...

AFAIU onnx runtime's support for WebGPU is still pretty minimal/experimental, so likely isn't able to run Whisper today Overview issue is here: https://github.com/microsoft/onnxruntime/issues/15796 There doesn't seem to be much up-to-date...

thanks for the update @guschmue ! Is there a GH issue for the problem you're describing? Is it this? https://github.com/microsoft/onnxruntime/issues/17373

@joefiorini just pointed me here: in order to fix compilation in [ember-cli](https://github.com/stefanpenner/ember-cli/issues/135), I just published a couple new broccoli plugins for compiling [LESS](https://github.com/gabrielgrant/broccoli-less-single) and [Stylus](https://github.com/gabrielgrant/broccoli-stylus-single) that are based (very) heavily...

Hey, thanks for the bug report. It's not immediately obvious to me what's breaking here. Did you ever figure this out?

Perhaps this is a silly question, but if the task is something performed on the model, is there a reason the task shouldn't live on the model?