fragro/llama-7b-hf

Model Overview

fragro/llama-7b-hf is a 7 billion parameter auto-regressive language model, part of the LLaMA family developed by Meta AI's FAIR team. Trained between December 2022 and February 2023, this model is a HuggingFace-compatible conversion of the original LLaMA-7B. It is based on the transformer architecture and was trained on a diverse dataset including CCNet, C4, GitHub, Wikipedia, and Books, totaling 1 trillion tokens.

Key Capabilities

• Research Foundation: Primarily intended for research on large language models, including exploring potential applications in question answering, natural language understanding, and reading comprehension.
• Bias Evaluation: Designed for evaluating and mitigating biases, risks, toxic content generation, and hallucinations inherent in language models.
• Multilingual Data: While predominantly English, the training data included content in 20 other languages, suggesting some multilingual understanding.

Intended Use Cases

• Academic Research: Ideal for researchers in natural language processing, machine learning, and artificial intelligence to study model capabilities and limitations.
• Foundation Model Exploration: Suitable for understanding the behavior of base language models before fine-tuning for specific downstream applications.

Limitations

As a foundational model, LLaMA-7B has not been trained with human feedback and can generate toxic or offensive content, incorrect information, or unhelpful answers. It is not intended for direct use in downstream applications without further risk evaluation and mitigation.