This would be cool. I suspect that people don't implement this model because it's a long tail of users requesting this. This is totally feasible technically.
Very cool. I'm a big fan of swipe. So much so that a bunch of us a Grammarly re-implemented it using a similar technique about 2 years ago on the Grammarly iOS keyboard, so if you want to experience something similar on iOS, you can head to the app store https://apps.apple.com/us/app/grammarly-ai-keyboard-notes/id... and use the qwerty english keyboard layout to get this neural net version of swipe.
One key difference is that the learned model does not decode in a context sensitive manner but does it a word at a time. The main reason is because we wanted to release this soon and wanted the user's personal dictionary (i.e. contact names, etc... to show up correctly when swiped). It would have been nice if we could have followed through with the context sensitive decoding as described by the FUTO folks. It would really help with accuracy when dealing with words like:
establishment_name: str
date: str = Field(description="YYYY-MM-DD")
total: float = Field(description="The total amount of the receipt")
currency: str = Field(description="The currency used for everything on the receipt")
items: list[Item] = Field(description="The items on the receipt")
</code>
There needs to be a better evaluation and a better provided schema that captures the full details of what is expected to be captured.
> What kind of error should it return if there's no total listed on the receipt? Should it even return an error or is it OK for it to return total = null?
Additionally, the schema allows optional fields, so the LLM is free to skip missing fields if they are specified as such.
I wonder if most people get the same things wrong. I checked your other comments and noticed that there is no server side component. In case you add one, I would be really interested in knowing which ones are most confused, and in which direction.
This is awesome! After about 50 attempts, I have a much better sense of what to look out for when I see something. I wish there were more such focused apps. for specific specific health related things.
How combustible are these batteries compared to the standard lower density ones, and if one of them catches fire, how easy/hard is it for the fire department to get it under control?
Facebook, PyTorch Mobile Team | US-Remote only | Full Time | Software Engineer (1 position available)
We’re the PyTorch Mobile Team (https://github.com/pytorch/pytorch/) working on making PyTorch broadly available for a plethora of mobile devices both within Facebook and outside. We’re interested in motivated engineers in this space who are willing to work remotely (within the US).
PyTorch is the most popular AI framework within the research community, and we’re working on making it production ready (especially on mobile devices).
You can see here (https://ai.facebook.com/tools/pytorch/) that the journey for Mobile (marked experimental) has just gotten started, and you’ll be jumping on to the bandwagon as it starts to leave the station!
Please drop us an email at (agaurav at fb dot com and dhruvbird at fb dot com), and we’d love to chat if this sounds interesting to you!
@kbenson Your explanation makes a lot more sense. Increasing the hash table size probably didn't affect the perf. significantly, but binding to the ANY ip did.
People try to make the fit be as tight as possible to the sample data -- the explanation is that simple. I don't buy the explanation provided in the article.
@bedhead hit the spot with his comment "Circumstances matter. I suppose if Uber hadn't entered into a partnership first, the whole thing would be more palatable. But by going this route, Uber is sending a clear signal to all potential future partners - watch out. I just find it distasteful, and sadly predictable given Uber's history."
Do you mean to say that it's okay for private operators to discrimate based on whether their potential client is disabled or not (when it has nothing to do with the safety of the client or the driver)?