Magnitude: Up to 2x Faster Than llama.cpp Officially, 21x Slower on Our CPU
Magnitude is an open-source inference engine for AI agents that auto-optimizes kernels for hardware, beating llama.cp ...
Ollama v0.35.0 Released: New API for Decision Models
Ollama v0.35.0 is released with a new/v1/systemone API for decision models, supporting nimble and tev1 models for fas ...
ComfyUI v0.38.0 Released: Hunyuan Image 3.5 and More
ComfyUI v0.38.0 is out, adding support for Hunyuan Image 3.5, ming-image, Qwen-Image 2.1 enhancements, w6a8 quantizat ...
unsloth Desktop v0.1.900-beta Released with Laya and Speedups
unsloth Desktop v0.1.900-beta adds local Laya execution, a Skills Editor, Library and Document Viewer, and major imag ...
InvokeAI v6.14.2 Released with Custom Fonts and Model Fixes
InvokeAI v6.14.2 is out, adding custom font imports on Canvas, style transfer for krea-2, VRAM management improvement ...
SmolDataEnvs: RL Tasks for Small Model Optimization
Explore SmolDataEnvs, a dataset with 5.5K+ RL tasks for hill-climbing optimization of small models in code and data s ...
KoboldCpp v1.122 Released with Built-in Agentic Framework
KoboldCpp v1.122 is released, featuring the new KoboldCpp Agent, ubatchsize controls, and upstream fixes from llama.c ...
RamaLama v0.25.0 Released: Security and Engine Updates
RamaLama v0.25.0 is out, featuring secure default bindings, llama.cpp v0.4.1 update, Vulkan support for NVIDIA, and W ...
ggml v0.25.2 Released with Hardware Backend Optimizations
ggml v0.25.2 has been released, focusing on improvements to hardware backends including CUDA, Vulkan, OpenCL, and Hex ...
llama.cpp v0.5.0 Released with Backend and Server Upgrades
llama.cpp v0.5.0 is released, featuring CUDA/Metal hardware acceleration, multi-address server support, and new model ...