One more time @github is being complicit of trying to censor @popcorntimetv. We've been through this before, you may like or not what that project is about but it is a GPL project and owns all of the IP on its code.
I hope @EFF and @fsf speak up
https://t.co/vz1FjlbFbX
Announcing Qwen2.5-VL Cookbooks!
🧑🍳A collection of notebooks showcasing use cases of Qwen2.5-VL, include local model and API. Examples include Compute use, Spatial Understanding, Document Parsing, Mobile Agent, OCR, Universal Recognition, Video Understanding.
🔗Link:https://t.co/iqwvTDrufT
💬 Qwen Chat: https://t.co/BhhXyzLt5B (choose Qwen2.5-VL-72B-Instruct as the model)
⚙️ API: https://t.co/o5wgAARK85
China is on fire, shipping non-stop! 🔥🔥
Hailuo just launched another insane Video model T2V-01-Director. It gives you full control over camera movement.
I had early access, and below are two of my favorite generations with their prompts:
NEW: DeepSeek Janus Pro 1B (Generate Images, Chat with PDF) running in your browser, 100% local, powered by WebGPU 🔥
Zero server costs, brought to you by transformers.js - try it out!
RIP Suno and Udio. 😬
China just dropped another open-source model: YuE, a full-song music generation model. From lyrics to complete songs, it can generate tracks up to minutes long.
It’s Hugging Face and LLAMA-compatible, for easy fine-tuning.
More details 👇
DeepSeek is on FIRE! 🔥 They just released Janus Pro: a multimodal LLM capable of visual understanding and image generation! 🤯
The 1B model can even run in your browser on WebGPU, powered by 🤗 Transformers.js!
This is the easiest way to run it locally: just visit a website!
There is an unprecedented level of cope around DeepSeek, and very little signal on X around R1. I recommend unfollowing anyone spreading conspiracy theories around R1/DeepSeek in general. (1/9)
Just in time for Christmas: a repository for decrypting many encrypted D-Link firmware images. Also integrated into Binwalk for auto-magic decryption & extraction.
https://t.co/4GcTrXoFhp
tomorrow, I'll be hosting a talk for @MIT; I'll be speaking about open-source computer vision tools
12:00 PM PST / 03:00 PM EST / 09:00 PM CET
we'll be streaming on X: https://t.co/aRhiw3zDq6
Announcing llama-ocr – a free + open source OCR tool!
It takes documents (images for now) & outputs markdown, and does really well for complex receipts, PDFs with tables/charts, ect...
Powered by Llama 3.2 vision on @togethercompute & available on npm today!
Here is my recent DEF CON talk on Anom, the encrypted phone secretly ran by the FBI. All about the phone, the network, how Anom was structured, who used it, what this means for Signal, Telegram, more https://t.co/IpSGqNk4UL
PSA i can spoof any https://t.co/Utqc6wxQRy email and it will pass all DKIM/SPF/etc. checks. here's an email i sent to myself pretending to be a famous MIT-affiliated podcaster - thanks gmail for auto-inserting the profile pic :)
MIT may fix this someday but in the meantime beware that it's trivial for any https://t.co/Utqc6wxQRy account to send mail as any other https://t.co/Utqc6wxQRy account!