The top 5 hugging face models measured by RiseFinder over the 1d window ending 13 September 2026.
A GGUF quantised build of a Qwen3.8 language model with its built in safety alignment deliberately removed, known as an abliterated or uncensored variant, packaged for use with the Unsloth fine tuning toolchain.
Hugging Face models
Open source ↗Instruction following and reasoning data prepared for supervised fine tuning.
Hugging Face models
Open source ↗Code problem solving data with the reasoning shown, split into worked solutions, direct answers and synthesised tasks.
Hugging Face models
Open source ↗Mathematical problem solving and dialogue data for training models to reason rather than to recall.
Hugging Face models
Open source ↗Google's time series forecasting model, pretrained so a series can be forecast without training anything first.
Hugging Face models
Open source ↗Get early signals on the websites, apps, repositories and startups gaining momentum. Join the list for Risefinder email briefings.
Be first to receive the briefing.