aliabd

Checkout the demo on spaces: https://huggingface.co/spaces/ysharma/Explore_llamav2_with_T...

Saw this on twitter: https://twitter.com/Aella_Girl/status/1645398628524064769

Super cool. Big fan of LLMs for explaining code/helping debug as opposed to autocomplete.

Have you tried playing with the prompt? Curious if something (so silly) like “how would an experienced engineer fix this error: <stacktrace>” make a difference.

Twitter thread by the journalist who confirmed details with the victims in the sample: https://twitter.com/_KarenHao/status/1543949945614393344

Just updated the link again - much faster now (inference down to 0.5s from 9s) Queue will move a lot quicker now.

They are the numbers at the bottom left of the interface under the 'clear' button. Try clicking them and it will auto fill the input then you just click submit. Or what do you see?

Just rebuilt the app with more examples, so the queue should stay a lot shorter now. This was built using Gradio[0] and is hosted on Spaces[1]. Check out the paper[2] and github repo[3].

[0]: https://github.com/gradio-app/gradio [1]: https://huggingface.co/spaces [2]: http://www.cvlibs.net/publications/Sauer2021NEURIPS.pdf [3]: https://github.com/autonomousvision/projected_gan

Thank you Nils! We were very fortunate to work with you and all of the awesome people at Pear!

We posted gradio as a Show HN mid 2020 [1] :) Got a lot of good feedback and have been slowly growing ever since!

[1]: https://news.ycombinator.com/item?id=23901834

Sorry we had to restart the interface - can you try again?

Sorry, we had to restart the interface for a separate issue. Please try again

Nothing is saved by default. Clicking 'flag' saves the input and output (only accessible to the space owner). More info on flagging: https://gradio.app/getting_started/#flagging

This was built on top of the Pytorch implementation [0] of AnimeGANv2, using gradio [1] and HuggingFace Spaces [2]. See the 100s of examples [3] posted by users.

[0]: https://github.com/bryandlee/animegan2-pytorch [1]: https://gradio.app [2]: https://huggingface.co/spaces [3]: https://twitter.com/ak92501/status/1457033482115420160/retwe...

You're right, see 100s of examples posted by users here: https://twitter.com/ak92501/status/1457033482115420160/retwe...

All within 439 days.

More info from creator: https://twitter.com/cyrildiagne/status/1450762117255204866

So I agree with you, but how would you formulate what makes text 'casual'?

My favorite:

[Formal] "Your mother is so old, her last name is asaurus" [Casual] "Yo mama's so old, her last name is asaurus"

[Casual] "Yo what's up with the weather?" [Formal] "What is the weather like?"

More info: https://github.com/PrithivirajDamodaran/Styleformer

Try sketching one of these: - bench (my favorite) - drums - airplane - basketball

This is awesome and will enable a whole suite of apps to be build on top of it. Great work

Sorry, this is because of the traffic right now. what you're seeing is a counter for how long the current prediction time is vs avg prediction time.

Yes the flag button is for wrong or unusual outputs. No worries though

Thanks for sharing! For context this is a demo of PaddleOCR V2 [0] which was released yesterday. You can find their original repo here [1]. We built this demo using Gradio [2] and deployed it on HuggingFace's Spaces [3].

[0]: https://arxiv.org/abs/2109.03144 [1]: https://github.com/PaddlePaddle/PaddleOCR [2]: https://gradio.app/ [3]: https://huggingface.co/spaces

Here's a web demo[0] where you can try out any two images and see the resulting hashes, and whether there's a collision. You can also try your own transformations (rotation, adding a filter, etc) on the image. Demo was built using Gradio[1].

[0]: https://huggingface.co/spaces/akhaliq/AppleNeuralHash2ONNX [1]: https://gradio.dev

I'm genuinely stunned that Robinhood would run over their #1 persona like this. I doubt they will ever recover

I think ML as a field just needs to really mature. A lot of the work feels super hacky, sort of a mix between research, demos, and production.

I'm the author here! Thanks for sharing :) Feel free to ask me any questions.

I wanted to highlight some of the libraries we used: Gradio [1] to create the interfaces in the GIFs, and PAIR-code's [2] Tensorflow implementation of the papers.

[1]: https://github.com/gradio-app/gradio [2]: https://github.com/PAIR-code/saliency

This is awesome! Love playing games like this with my parents (long distance), but they only ever want to play on their phones and the UI is always terrible. Wish I could've tried some of the games though without needing someone else to join. Some sort of automated example or a general open room?

(Reposting my comment here from the other thread)

AFAIK, reverse engineering APIs isn’t illegal. [1]

There are some (VC backed) start ups that are open about doing it. Teller API is an example. [2]

[1] “Reverse engineering generally doesn't violate trade secret law because it is a fair and independent means of learning information, not a misappropriation. Once the information is discovered in a fair and honest way, it also can be reported without violating trade secret law.“ https://www.eff.org/issues/coders/reverse-engineering-faq [2] “We reverse engineer these apps to discover their secret API contracts and then implement clients for them.” https://teller.io/

AFAIK, reverse engineering APIs isn’t illegal. [1]

There are some (VC backed) start ups that are open about doing it. Teller API is an example. [2]

[1] “Reverse engineering generally doesn't violate trade secret law because it is a fair and independent means of learning information, not a misappropriation. Once the information is discovered in a fair and honest way, it also can be reported without violating trade secret law.“ https://www.eff.org/issues/coders/reverse-engineering-faq

[2] “We reverse engineer these apps to discover their secret API contracts and then implement clients for them.” https://teller.io/