Gemma3-1B-ANE

Gemma3-1B for Apple ANE

Quickstart

1. Install NexaSDK

2. Run the model with one line of code:

   nexa infer NexaAI/gemma3-1b-ane
   

Model Description

Gemma3-1B is a lightweight 1-billion-parameter language model developed by Google as part of the Gemma 3 series.
Designed for on-device and resource-constrained environments, it delivers strong language understanding, fast inference, and energy-efficient deployment.

Despite its compact size, Gemma3-1B supports broad NLP tasks and can be fine-tuned for both general and domain-specific applications.

Features

• Ultra-lightweight: optimizes for fast, low-memory inference.
• Conversational AI: maintains coherent, context-aware dialogue.
• Text generation: summaries, drafts, comments, and structured content.
• Efficient reasoning: small-scale analytical and step-by-step outputs.
• Multilingual: handles multiple languages within its training scope.
• Highly portable: ideal for edge devices and on-device applications.

Use Cases

• Mobile and embedded AI assistants
• On-device chatbots with strict latency/power limits
• Local summarization for apps (notes, email, notifications)
• Lightweight agents for IoT, robotics, and wearables
• Fine-tuned task-specific models (classification, command execution, etc.)

Inputs and Outputs

Input:

• Text prompts or conversation history (tokenized sequences for model APIs).

Output:

• Generated text (answers, explanations, creative output).
• Optional logits/probabilities for downstream tasks.

License

This repo is licensed under the Creative Commons Attribution–NonCommercial 4.0 (CC BY-NC 4.0) license, which allows use, sharing, and modification only for non-commercial purposes with proper attribution. All NPU-related models, runtimes, and code in this project are protected under this non-commercial license and cannot be used in any commercial or revenue-generating applications. Commercial licensing or enterprise usage requires a separate agreement. For inquiries, please contact [email protected]