llama.cpp official releases
Llama.app Server Fixes Draft Context Memory Allocation
Why is the server suddenly returning 500 errors? A draft context update might hold the answer.
The complete AI news desk
The AI world, in one verified news room.
Models, agents, robotics, research, business and policy, with the original source attached. Search and move beyond the latest thirty updates.
Latest and archive
Open the fast flashes streamInfrastructure
llama.cpp official releases
SYCL Kernel Changes Reverted in Latest LLaMA.cpp Update
Why the sudden reversal in kernel commits?
Ollama releases
New Release Candidate Introduces Model Recommendation Feature
Unlock the future of recommendations with the latest update. What's new?
llama.cpp official releases
b10548: Corrects Typographical Errors in ET.md
Typo troubles: What's changed in ET.md?
llama.cpp official releases
Tensor API Fixes Out-of-Bounds Issue in Metal Kernel
Why do some tensor operations yield NaNs on macOS?
llama.cpp official releases
Nvidia Enhances CUDA for Faster Quantized Decode on RTX 5090
Can NVIDIA's latest tweaks boost your GPU's performance?
llama.cpp official releases
Metal Backend Enhances Flash Attention with Dequantization Update
Why is the Metal backend transforming quantized data before flash attention?
llama.cpp official releases
Llama.app Reverses Recent Tensor-Split Meta Backend Changes
What's behind the sudden reversal of the "tensor-split" update?
llama.cpp official releases
IBM Engineer Commits Extensive Updates to GraniteSWAForCausalLM
What's behind IBM's recent flurry of AI model updates?
llama.cpp official releases
OpenCL Project Commits to Deterministic Expert Scatter in MoE
Unpredictable code behavior may soon be tamed.
Ollama releases
Ollama's v0.32.15 Enhances Efficiency with Metadata Cache
Discover the secret behind Ollama's latest speed boost.
Ollama releases
mlx-c#127 Temporarily Integrated in v0.32.15-rc0
Unlock the mystery behind the latest mlx update in v0.32.15-rc0. What's in store?
llama.cpp official releases
Router Command Messages Suppressed in Latest Update
Mysterious server messages vanish: What's behind the silence?
llama.cpp official releases
Llama.app Enhances Build System with CMake Clean-up and Target Aliases
Why are developers switching to namespaced aliases?
llama.cpp official releases
NVIDIA Adjusts MMVQ Parameters for Dense Models on DGX Spark
Nvidia's latest CUDA tweaks spark curiosity. What's changing for dense models?
llama.cpp official releases
AMD Updates HIP Builds for Accurate Memory Reporting on APUs
Why are AMD APUs suddenly revealing more memory?
llama.cpp official releases
Intel's SYCL Kernel Optimization Boosts Arc 70 Performance
Why did the q4_0 path see a massive speed boost on Arc 70?
Ollama releases
Ollama Updates WebP Support and Message Handling
What's next for llama-server? Discover the latest updates in v0.32.14.
You reached the end of the available verified archive.