Pulsar Rust + CUDA engine that streams giant MoE models from SSD, running models far larger than VRAM. Mixed-precision GGUFs. github.com/giannisanni/pulsar Collection by giannisan 1 day ago - giannisan/Hy3-ds4-gguf 299B • Updated 9 days ago • 18.8k • 29 giannisan/GLM-5.2-ds4-gguf 754B • Updated 17 days ago • 1.17k • 3 giannisan/Laguna-S-2.1-ds4-gguf 118B • Updated 1 day ago • 20 • 2
BIG Collection by gsting 11 days ago - deepseek-ai/DeepSeek-V4-Pro Text Generation • 862B • Updated Jun 22 • 1.62M • • 5.3k meituan-longcat/LongCat-2.0 Text Generation • 1.8T • Updated 16 days ago • 2.68k • 216 MiniMaxAI/MiniMax-M3 Image-Text-to-Text • 427B • Updated about 23 hours ago • 178k • • 1.36k mistralai/Mistral-Large-3-675B-Instruct-2512 Updated 9 days ago • 847 • 239
MLX Collection by mfxler 8 days ago - mlx-community/MOSS-SoundEffect-MLX-4bit Text-to-Audio • Updated May 27 • 3 utter-project/EuroLLM-1.7B-Instruct Text Generation • 2B • Updated Dec 16, 2024 • 8.62k • 101 jasonvassallo/mlx-audio-generate Text-to-Audio • Updated Feb 25 • 2 aufklarer/MAGNeT-Small-30secs-MLX-4bit Text-to-Audio • 80.6M • Updated May 20 • 45
AI Collection by azore89 14 days ago - k2-fsa/OmniVoice Text-to-Speech • 0.6B • Updated 21 days ago • 831k • 1.19k
VLADrop: Drop-Then-Recovery (DTR) Checkpoints Checkpoints for 'Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?' (arXiv:2606.27755). Code: https://github.com/s1ghhh/VLADrop Collection by s1ghhh 11 days ago - s1ghhh/VLADrop_pi05_LIBERO_Baseline Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop9Block_Uniform Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop16Block_Endpoints Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop9Block_GateProbe_FLOPsMatched Robotics • 4B • Updated 14 days ago
Speech Collection by tyty-cli-guy 14 days ago - Running Featured 461 HF Realtime Voice 🎙 461 Voice chat over WebSocket against a HF speech-to-speech
text-to-action Collection by Mio99 14 days ago - Running on Zero Agents Featured 293 HY-Motion-1.0 💃 293 Text-to-3D and Image-to-3D Generation
PixelDiT Collection by Lextop 14 days ago - Comfy-Org/PixelDiT Updated 7 days ago • 76.5k • 133 nvidia/PiD Image-to-Image • Updated 14 days ago • 255 • 393
LTE Learning from Trial and Error. Collection by PKU-Wu-Lab 14 days ago - PKU-Wu-Lab/LTE-Qwen3-4B-Base Text Generation • 4B • Updated 14 days ago • 377 PKU-Wu-Lab/LTE-Qwen3-8B-Base Text Generation • 8B • Updated 14 days ago • 372 Do Not Step Into the Same River Twice: Learning to Reason from Trial and Error Paper • 2510.26109 • Published Jan 6
Do Not Step Into the Same River Twice: Learning to Reason from Trial and Error Paper • 2510.26109 • Published Jan 6
Pulsar Rust + CUDA engine that streams giant MoE models from SSD, running models far larger than VRAM. Mixed-precision GGUFs. github.com/giannisanni/pulsar Collection by giannisan 1 day ago - giannisan/Hy3-ds4-gguf 299B • Updated 9 days ago • 18.8k • 29 giannisan/GLM-5.2-ds4-gguf 754B • Updated 17 days ago • 1.17k • 3 giannisan/Laguna-S-2.1-ds4-gguf 118B • Updated 1 day ago • 20 • 2
VLADrop: Drop-Then-Recovery (DTR) Checkpoints Checkpoints for 'Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?' (arXiv:2606.27755). Code: https://github.com/s1ghhh/VLADrop Collection by s1ghhh 11 days ago - s1ghhh/VLADrop_pi05_LIBERO_Baseline Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop9Block_Uniform Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop16Block_Endpoints Robotics • 4B • Updated 14 days ago s1ghhh/VLADrop_pi05_LIBERO_Drop9Block_GateProbe_FLOPsMatched Robotics • 4B • Updated 14 days ago
BIG Collection by gsting 11 days ago - deepseek-ai/DeepSeek-V4-Pro Text Generation • 862B • Updated Jun 22 • 1.62M • • 5.3k meituan-longcat/LongCat-2.0 Text Generation • 1.8T • Updated 16 days ago • 2.68k • 216 MiniMaxAI/MiniMax-M3 Image-Text-to-Text • 427B • Updated about 23 hours ago • 178k • • 1.36k mistralai/Mistral-Large-3-675B-Instruct-2512 Updated 9 days ago • 847 • 239
Speech Collection by tyty-cli-guy 14 days ago - Running Featured 461 HF Realtime Voice 🎙 461 Voice chat over WebSocket against a HF speech-to-speech
MLX Collection by mfxler 8 days ago - mlx-community/MOSS-SoundEffect-MLX-4bit Text-to-Audio • Updated May 27 • 3 utter-project/EuroLLM-1.7B-Instruct Text Generation • 2B • Updated Dec 16, 2024 • 8.62k • 101 jasonvassallo/mlx-audio-generate Text-to-Audio • Updated Feb 25 • 2 aufklarer/MAGNeT-Small-30secs-MLX-4bit Text-to-Audio • 80.6M • Updated May 20 • 45
text-to-action Collection by Mio99 14 days ago - Running on Zero Agents Featured 293 HY-Motion-1.0 💃 293 Text-to-3D and Image-to-3D Generation
AI Collection by azore89 14 days ago - k2-fsa/OmniVoice Text-to-Speech • 0.6B • Updated 21 days ago • 831k • 1.19k
PixelDiT Collection by Lextop 14 days ago - Comfy-Org/PixelDiT Updated 7 days ago • 76.5k • 133 nvidia/PiD Image-to-Image • Updated 14 days ago • 255 • 393
LTE Learning from Trial and Error. Collection by PKU-Wu-Lab 14 days ago - PKU-Wu-Lab/LTE-Qwen3-4B-Base Text Generation • 4B • Updated 14 days ago • 377 PKU-Wu-Lab/LTE-Qwen3-8B-Base Text Generation • 8B • Updated 14 days ago • 372 Do Not Step Into the Same River Twice: Learning to Reason from Trial and Error Paper • 2510.26109 • Published Jan 6
Do Not Step Into the Same River Twice: Learning to Reason from Trial and Error Paper • 2510.26109 • Published Jan 6