Official Python inference and LoRA trainer package
The most powerful local music generation model
Qwen3-Coder is the code version of Qwen3
Video Object and Interaction Deletion
Programmatic access to the AlphaGenome model
A GPT-4o Level MLLM for Vision, Speech and Multimodal Live Streaming
A 0.1B Omni model trained from scratch
Tool for exploring and debugging transformer model behaviors
CLIP, Predict the most relevant text snippet given an image
Official DeiT repository
Open-source pre-training implementation of Google's LaMDA in PyTorch