Running Featured 1.4k FineWeb: decanting the web for the finest text data at scale π· 1.4k Explore and download the FineWeb webβscale text dataset
Running on CPU Upgrade Featured 3.24k The Smol Training Playbook π 3.24k The secrets to building world-class LLMs
deepseek-ai/DeepSeek-V3.2-Speciale Text Generation β’ 685B β’ Updated Dec 1, 2025 β’ 1.01k β’ β’ 716
nvidia/Llama-3_1-Nemotron-Ultra-253B-v1 Text Generation β’ 253B β’ Updated Oct 15, 2025 β’ 2.69k β’ 352
meta-llama/Llama-4-Scout-17B-16E-Instruct Image-Text-to-Text β’ 109B β’ Updated May 22, 2025 β’ 462k β’ β’ 1.33k