Smollm2 - > quen changes fix for deploy at hf2- fix to docker

Smollm2 - > quen changes fix for deploy at hf2

Smollm2 - > quen changes fix for deploy at hf

Smollm2 - > quen changes

whisper<-vosk2 changes

whisper<-vosk2

whisper->vosk2

whisper->vosk

Simplify API to single prompt -> single output

change in task

Switch to faster-whisper with small model

Switch to whisper base model for better accuracy

Pin numpy<2 for torch 2.2.2 compatibility

Fix pkg_resources build error

Fix pkg_resources build error

Add nlp-asr-kit API

initial commit