• HN Mail
  • Subscribe

ML

Unsolved Problems in MLOps
52 points | 5 comments

MLB restricts using dugout iPads for AI-assisted in-game strategy
6 points | 0 comments

A new ML compiler to run 70B+ LLMs on consumer GPUs with <1% accuracy loss
4 points | 3 comments

6.4x faster than llama.cpp, 3.9x faster than MLX
4 points | 2 comments

BioChain: A Layer-1 blockchain signed with ML-DSA-44 (post-quantum) url
3 points | 0 comments

Crystal ML Library: Autograd, Tensors, Neural Networks, Optimizers
3 points | 0 comments

Hedos – Local models in one native Mac app (llama.cpp, MLX, diffusers)
3 points | 0 comments