Atomic-Germ / Llama-3.1-8B-Instruct-NPU2

🦙 LLaMA 3.1 (8B) – Optimized for FastFlowLM on AMD Ryzen™ AI NPU (XDNA2 Only)

Model Summary

This is a derivative of Meta AI’s LLaMA 3.1 base model. The model retains the core architecture and weights from Meta’s release and may include fine-tuning, quantization, or adaptation for specific applications.

⚠️ This model is subject to Meta’s LLaMA 3 license. You must accept Meta’s terms to use or download it.

📝 License & Usage Terms

Meta LLaMA 3 License

• Base model is governed by Meta AI's license:
  👉 https://ai.meta.com/llama/license/

• You must agree to their license terms to access and use the weights, which include:

  • No commercial use without permission
  • Redistribution only allowed under specific conditions
  • Attribution required

Redistribution Notice

• This repository does not include original Meta weights.
• You must obtain base weights directly from Meta:
  👉 https://huggingface.co/meta-llama

If Fine-tuned

If this model has been fine-tuned, the downstream weights are provided under the following conditions:

• Base Model License : Meta’s LLaMA 3 License
• Derivative Weights License : [e.g., CC-BY-NC-4.0, MIT, custom, etc.]
• Training Dataset License(s) :
  • [Dataset A] – [license]
  • [Dataset B] – [license]

Make sure you have rights to use and distribute any data used in fine-tuning.

Intended Use

• Use Cases : Research, experimentation, academic NLP, code generation (if applicable)
• Not Intended For : Use in production systems without further evaluation, sensitive applications, or commercial deployments without Meta’s explicit permission

Limitations & Risks

• May generate incorrect or harmful content
• Does not have knowledge past its training cutoff
• Biases in training data may persist

Citation

@misc{touvron2024llama3,
  title={LLaMA 3: Open Foundation and Instruction Models},
  author={Touvron, Hugo and others},
  year={2024},
  url={https://ai.meta.com/llama/}
}