I thank Microsoft deeply for the forced copilot crap, almost impossible to remove, that they have put into vs code. Finally after 5 years I have deleted vs code in my Mac! that was the last piece of windows software I still had around. VS code was great years ago, until Microsoft started to push crap into it – and afaik they also made the fully open-source, telemetry-free fork difficult to use with many extensions.
Really, thanks for forcing me into deleting it. turns out vim + Claude Code or codex was much better all along, it really works well for me.
anything that doesn't touch the model parameters at all once it has been compiled. for example, in streaming ASR of an encoder-decoder you can get gains in accuracy just by enhancing the encoder-decoder orchestration and ratio, frequency of fwd passes, dynamically adjusting the length of rolling windows (if using full attention). Prompting would be part of this too, including few-shot examples. Decoding strategy is also part of this (top-k, nucleus, speculative decoding, greedy or anything else). Applying signal processing or any kind of processing to the input before getting it into the model, or to the output. There are a lot of things you can do.
we constantly underestimate the power of inference scaffolding. I have seen it in all domains: coding, ASR, ARC-AGI benchmarks you name it. Scaffolding can do a lot! And post-training too. I am confident our currently pre-trained models can beat this benchmark over 80% with the right post-training and scaffolding. That being said I don't think ARC-AGI proves much. It is not a useful task at all in the wild. it is just a game; a strange and confusing one. For me this is just a pointless pseudo-academic exercise. Good to have, but by no means measures intelligence and even less utility of a model.
I am not sure what "edge" device you want to run this on, but you can compress parakeet to under 500MB on RAM / disk with dynamic quants on-the-fly dequantization (GGUF or CoreML centroid palettization style). And retain essentially almost all accuracy.
And just to be clear, 500MB is even enough for a raspberry Pi. Then your problem is not memory, is FLOPS. It might run real-time in a RPi 5, since it has around 50 GFLOPS of FP32, i.e. 100 GFLOPS of FP16. So about 20-50 times less than a modern iPhone. I don't think it will be able to keep it real time, TBF, but close.
regardless, this model with such quantization strategy runs real time at +10x real-time factor even in 6-year old iPhones (which you can acquire for under $200) and offline at a reasonable speed, essentially anywhere.
You get the best of both worlds: the accuracy of a whisper transformer at the speed and footprint of a small model.