Hey everyone, super exicted about the qwen release and thus I created a demo to try out my uncensored version. Check it my space here: https://huggingface.co/spaces/JonathanColetti/Qwen3.8-27B-Uncensored-Demo / https://huggingface.co/JonathanColetti/Qwen3.8-27B-Uncensored-GGUF
Hi everyone, I spent a lot of resources (renting gpu clusters) to train something similar to wan streamer (on a smaller scale). If you could check it out that would be awesome. I tried to make it as small as possible but realtime inference is expensive. Check it out on github: https://github.com/JonathanColetti/LiveWan or on huggingface: https://huggingface.co/JonathanColetti/LiveWan thank you! (wan2.1 only supports 5 seconds of video and the women talking is around 4mins)
No has been built a replica of the architecture of deepseek v3.2, so I did it.
You can check it out here thanks: https://github.com/JonathanColetti/deepseek-v3.2-pytorch