Kai-Fu Li's Yi-34B uses exactly Llama's architecture except for 2 tensor renamed

MistralLite by Amazon Web Services

Real-Time Latent Consistency Model

Sparse LLM Inference on CPU: 75% fewer parameters

Segmind Stable Diffusion – A smaller version of Stable Diffusion XL

Zephyr 7B – Mistral Finetune that responds like ChatGPT

Show HN: MiniSearch, a minimalist search engine with integrated browser-based AI

Accelerating Stable Diffusion XL Inference with Jax on Cloud TPU v5e

Mistral-7B-OpenOrca. First 7B model to beat all other models <30B

PubMedBERT Embeddings - Semantic search and retrieval augmented generation for medical literature