1sec.ai

Tag

#edge-ai

Every item tagged edge-ai, newest first.

6 items

models9h

I released Inflect-Nano, an ultra-extreme tiny 4.63m parameter TTS model.

The Inflect-Nano-v1 TTS model has 4.63m parameters, making it the second-smallest publicly released TTS model after TinyTTS. Despite its tiny size, it reportedly performs well for its model weight. The model can run on very low-end hardware, making it suitable for deployment on resource-constrained devices. You can experiment with this model for edge cases where compute is extremely limited.

Key takeaways
  • 4.63m parameters, second-smallest public TTS model.
  • Runs on very low-end hardware, suitable for resource-constrained devices.
  • Not state-of-the-art, but functional for its size.
rr/LocalLLaMA#tiny-llm#tts#edge-ai
modelsMar 7

LLM Inference on Edge: A Fun and Easy Guide to run LLMs via React Native on your Phone!

The guide walks you through running LLMs on mobile devices using React Native, leveraging Hugging Face's Transformers.js library for efficient inference. This enables developers to deploy LLMs in mobile apps, expanding AI accessibility. The approach focuses on optimizing performance for edge devices. You can explore code examples and demos to get started.

Key takeaways
  • LLMs can run on mobile devices via React Native.
  • Hugging Face's Transformers.js library enables efficient inference.
  • Optimized for edge devices, expanding AI accessibility.
HHugging Face Blog#edge-ai#mobile-llm#react-native
modelsSep 25

Llama can now see and run on your device - welcome Llama 3.2

Meta released Llama 3.1, an update to the Llama model family that adds on-device execution capabilities. The model can run locally on devices with sufficient RAM, expanding deployment options for builders. Local execution enables lower latency and no reliance on cloud infrastructure. This update targets applications requiring real-time responses or offline functionality.

Key takeaways
  • Llama 3.1 supports on-device execution on devices with sufficient RAM.
  • Enables lower latency and offline functionality.
  • Expands deployment options for local and edge applications.
HHugging Face Blog#local-llm#on-device#edge-ai
modelsJul 16

SmolLM - blazingly fast and remarkably powerful

SmolLM is a new open-source language model that prioritizes speed and efficiency. It achieves state-of-the-art performance on several benchmarks while requiring significantly less computational resources. The model is designed for deployment on edge devices and other resource-constrained environments. You can access SmolLM through the Hugging Face model hub.

Key takeaways
  • SmolLM achieves state-of-the-art performance with less computational resources.
  • Designed for deployment on edge devices and resource-constrained environments.
  • Available on the Hugging Face model hub.
HHugging Face Blog#open-source#edge-ai#efficient-models
otherApr 2

Bringing serverless GPU inference to Hugging Face users

Cloudflare partners with Hugging Face to enable serverless GPU inference on the Hugging Face platform. This integration allows users to deploy and run AI models at the edge, reducing latency and costs. Builders can now access GPU-accelerated inference without managing infrastructure. The partnership aims to make AI model deployment more accessible and efficient.

Key takeaways
  • Serverless GPU inference now available on Hugging Face via Cloudflare.
  • Deploy AI models at the edge to reduce latency and costs.
  • No infrastructure management required for users.
HHugging Face Blog#serverless#gpu-inference#edge-ai
otherFeb 14

AMD Pervasive AI Developer Contest!

AMD is running a developer contest focused on building AI applications with Hugging Face models. The contest aims to encourage innovation in AI development, with a specific emphasis on edge AI and pervasive AI. Developers can participate by submitting their projects, with the opportunity to win prizes. This contest is relevant to builders looking to leverage AI at the edge.

Key takeaways
  • The contest focuses on edge AI and pervasive AI applications.
  • Developers can use Hugging Face models for their projects.
  • There are prizes available for the winners.
HHugging Face Blog#ai-contest#edge-ai#hugging-face