- Did you train the encoder & decoder together or separately? It would be nice to have the encoder representation be compatible with the existing whisper implementation since it would mean you could swap your implementation into models where its used as a component, like in the recent Voxtral model. I'd imagine it also might make training a bit faster as well.

- Did you consider training the turbo model as well?

Here(https://github.com/openai/whisper/blob/main/whisper/model.py...) is the relevant code in the whisper repo. You'd just need to change the for loop to an enumerate and subsample the context along its length at the point you want. I believe it would be:

for i, block in enumerate(self.blocks): x = block(x) if i==4: x = x[,,::2]

I'm not against using simple methods where appropriate. 95% of the companies out there probably do not need frameworks like spark. I think the main argument against them is operational complexity though, not the compute overhead.

Imagine an example where you have two arrays of 1 billion numbers, and you want to add them pairwise. You could use spark to do that by having each "task" be a single addition. But the time it would take to structure and transmit the 1 billion requests will be many multiples of the amount of time it would take to just do the additions.

Individual tasks would definitely have better latency, but I'd suspect the impact on throughput/CPU usage might be muted. Of course at the extremes (very small jobs, very large/complex objects being passed) you'd see big gains.

It would be great to see programs like this in the US. The closest I can think of is MealPal, where they simplify logistics by having each restaurant only focus on 1-2 dishes so they can churn them out incredibly quickly. I know Marc Lore's Wonder is also trying something similar with ghost kitchens, but don't know much of the details. Being able to serve more people with the same store space also cuts down on the fixed costs/person. I enjoy cooking, but it is hugely inefficient for everyone to do it for themselves at a civilizational level. If we could make healthy restaurants cost competitive with home cooking, it would probably equate to trillions in saved time.

[1] https://www.ssb.no/en/statbank/table/08313

Ironically, it's felt like US tech companies have been moving in the opposite direction lately. There seems to be a lot of demand for very senior engineers at the same time it's become increasingly difficult for new grads to find jobs. Maybe this is just a function of the pinch in CS graduation rates ~15 years ago[1]. From my experience as a manager, the premium you pay for senior talent also seems like a good deal. Even among good programs, there is huge variance in the productivity of new grads and training is a difficult proposition when changing jobs every 2 years is common.

[1] https://www.jamiefosterscience.com/wp-content/uploads/2023/1...

https://www.firstpost.com/tech/tech-layoffs-2024-dell-fires-...