Trending

Show HN: We made our own inference engine for Apple Silicon

Grok 4 Heavy ($300/mo) returns its surname and no other text: "Hitler"

Grok 4 Heavy Protects it's System prompt

LLM Inference Handbook

FP8 is ~100 tflops faster when the kernel name has "cutlass" in it

Regarding Prollyferation: Followup to "People Keep Inventing Prolly Trees"

Grok 4 Launch [video]

Launch HN: Morph (YC S23) – Apply AI code edits at 4,500 tokens/sec

BharatMLStack – Realtime Inference, MLOps

Totally lightweight local inference...

[R] Interactive Probabilistic Neural Network Decision Matrix Model

support for Kimi-K2 has been merged into llama.cpp

EXAONE 4.0 32B

[R] Unlearning Comparator — A Visual Analytics Toolkit for Machine Unlearning

MLB random forest with 53%-60% training accuracy. Prediction probability question. [P]

"We will release o3 wieghts next week"

Support for the upcoming IBM Granite 4.0 has been merged into llama.cpp

Grok 4 is dropped Guys!!! 🔥🔥😆

microsoft

Save item

onnxruntime

ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

huggingface

Save item

pytorch-image-models

The largest collection of PyTorch image encoders / backbones. Including train, eval, inference, export scripts, and pretrained weights -- ResNet, ResNeXT, EfficientNet, NFNet, Vision Transformer (ViT), MobileNetV4, MobileNet-V3 & V2, RegNet, DPN, CSPNet, Swin Transformer, MaxViT, CoAtNet, ConvNeXt, and more

modelscope

Save item

modelscope

ModelScope: bring the notion of Model-as-a-Service to life.