LLaMA: Open and Efficient Foundation Language Models
Meta released LLaMA, a collection of foundation language models ranging from 7B to 65B parameters. These models match larger proprietary models while being more efficient.
Brief
Ranked AI/ML cut for builders
High-Resolution Image Synthesis with Latent Diffusion Models
Latent diffusion models enable high-resolution image synthesis by applying diffusion in latent space. This approach dramatically reduces computational requirements.
Language Models are Few-Shot Learners
GPT-3 demonstrated that scaling language models improves task-agnostic few-shot performance. The 175B parameter model achieved strong results without fine-tuning.
BERT: Pre-training of Deep Bidirectional Transformers
Bidirectional encoder representations from transformers enable fine-tuning for many NLP tasks. BERT set new benchmarks across eleven natural language processing tasks.
Attention Is All You Need
The transformer architecture paper that introduced self-attention mechanisms for sequence transduction. This foundational work eliminated recurrence and convolutions entirely.