April 1st is just a day for unfunny people to be unfunny but louder. =_=
https://nixsanctuary.com/richard-stallman-has-finally-given-windows-the-green-light-as-a-desktop-operating-system/
https://nixsanctuary.com/richard-stallman-has-finally-given-windows-the-green-light-as-a-desktop-operating-system/
The flow is still not AS SMOOTH as the translation service you can already set up with #Retroarch in my opinion. But I feel I can get there with some more work.
So more progress on my project. It actually properly waits for the screenshot to finish being made before trying to upload it. Aside from the one image that after two tries just did not want to translate shows more promise. Was using `qwen2.5-vl-7b-instruct` via their api but can easily point to llama.cpp like setup once the new qwen2.5-vl models can be hosted via llama-serve also would probably be faster as well.
(Sorry for leaving cursor on the game during the recording
)
(Game is Doki Doki Pretty League for the PlayStation)
(Sorry for leaving cursor on the game during the recording

(Game is Doki Doki Pretty League for the PlayStation)
@jorgaborg I wish it was that simple. I just want a reason to play the game again there is no other puzzle game to my knowledge like it in terms of game play.
Man I wish #Catherine Full Body would get a PC release.

Progress on my dumb project to make fumbling through Japanese games a little easier. I think I could really get better results with a a more fleshed out prompt and I need to wait like half a second or so once the new image is in the folder and we might be getting somewhere.
(Oh and the game in the video is Urusei Yatsura: Stay With You for the PCE-CD. Just picked a random game to test.)
(Oh and the game in the video is Urusei Yatsura: Stay With You for the PCE-CD. Just picked a random game to test.)

I need to work on the prompt better as it could format its output better & the translations I am not sure yet if better to send off somewhere else for translation compared to letting the LLM do it but overall I am extremely impressed. I might have to make a actual project out of this since Qwen2.5-Omni-7B has so much power for such a small model.
Qwen2.5-Omni-7B just released and it is easily on one my new favorite LLMs as it can take Text, Audio, Images, & Video as input and even gives you audio as a output with the text.
https://huggingface.co/Qwen/Qwen2.5-Omni-7B
You can play with it here:
https://huggingface.co/spaces/Qwen/Qwen2.5-Omni-7B-Demo
https://huggingface.co/Qwen/Qwen2.5-Omni-7B
You can play with it here:
https://huggingface.co/spaces/Qwen/Qwen2.5-Omni-7B-Demo

I got around to really playing Slay the Princess and...maybe I am old and burn out but man it tries to hard and I got really bored. Like I feel it is just "so deep" normie bait. Which kinda just leaves me feeling empty and kinda wish I never engaged with it at all.
Slightly better #RealESRGAN results with the #PSX staring resolution of 320x224.
I wonder if there is something better I can be using than #RealESRGAN to try an upscale this or am I just bad at using the tools. Could also be that the source video is just real hard to scale since comes from a #NintendoDS cart with a resolution of 256x192.