#huggingface

• LFM2.5-2.6B: Local agents with 128K context, no cloud required

  LiquidAI's LFM2.5-2.6B runs agentic workflows on-device with 128K context, tool calling, and an OpenAI-compatible endpoint. No cloud inference bill.

  news · models · datasets · open-source

  Aug 4 · 2 min

• 🤗 Kernels: Signing, Trusted Publishers, and Hub Integration

  Hugging Face revamped Kernels with cryptographic signing, trusted publisher verification, and first-class Hub integration. Here's what changed.

  news · models · datasets · open-source

  Jul 6 · 2 min

• Photoroom PRX: Inside Their Training Data Strategy

  Photoroom's PRX Part 4 details their training data strategy — sourcing, curation, synthetic data, and open dataset releases on HuggingFace.

  news · models · datasets · open-source

  Jul 6 · 2 min

• Hugging Face + Cerebras: Gemma 4 now runs speech-to-speech in real time

  Hugging Face and Cerebras built a real-time voice AI pipeline using Google's Gemma 4 model, cutting latency to conversational speeds with open-source components.

  news · models · datasets · open-source

  Jul 1 · 2 min