EryriLabs/llama-3.2-uk-legislation-3b

Llama 3.2 UK Legislation 3B

This model is the base version of Meta's Llama 3.2 3B architecture. It has been pretrained on UK legislative texts but has not yet been fine-tuned. It is provided as a foundational model for further development, such as domain-specific applications or fine-tuning for specialised tasks involving UK legislative documents. It was trained as part of a blog series, see the article here

Model Details

Model Description

• Developed by: GPT-LABS.AI
• Model type: Transformer-based language model
• Language: English
• License: CC BY 4.0
• Base model: unsloth/Llama-3.2-3B

Model Sources

• Repository: EryriLabs/llama-3.2-uk-legislation-3b
• Blog Post: Making a Domain-Specific UK Legislation LLM: Part 1 - Pretraining

Uses

Intended Use

This base model is designed to serve as a starting point for further fine-tuning and development for tasks such as:

• Domain-specific applications in law or other fields
• Research and experimentation in natural language processing
• General-purpose natural language understanding and generation

Out-of-Scope Use

This model is not suitable for:

• Providing domain-specific expertise or insights without fine-tuning
• Applications requiring high accuracy or nuanced understanding of UK legislation
• Tasks involving sensitive or critical real-world applications without rigorous evaluation

Bias, Risks, and Limitations

• Bias: The model may reflect biases inherent in the pretraining data. Outputs should be critically evaluated for accuracy and fairness.
• Risks: As a base model, it may generate responses that are overly general or contextually inappropriate for specific tasks.
• Limitations: The model is not fine-tuned for specific domains, including legal or legislative text, and does not include the most recent developments in any field.

How to Get Started with the Model

from transformers import AutoModelForCausalLM, AutoTokenizer

# Load model and tokenizer
model = AutoModelForCausalLM.from_pretrained("EryriLabs/llama-3.2-uk-legislation-3b", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained("EryriLabs/llama-3.2-uk-legislation-3b")

# Sample question
input_text = "What are the main principles of UK legislation?"

# Tokenize and generate response
inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
outputs = model.generate(inputs["input_ids"], max_length=50)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)

print(response)

Technical Specifications

• Model Architecture: Llama 3.2 3B, a transformer-based model designed for natural language processing tasks.
• Training Data: Pretrained on a diverse dataset of general text.
• Compute Infrastructure: Training conducted on high-performance GPUs (e.g., NVIDIA A100).

Citation

If you use this model, please cite:

@misc{llama3.2-uk-legislation-3b,
  author = {GPT-LABS.AI},
  title = {Llama 3.2 UK Legislation 3B},
  year = {2024},
  publisher = {Hugging Face},
  url = {https://huggingface.co/EryriLabs/llama-3.2-uk-legislation-3b}
}

Model Card Authors

• GPT-LABS.AI

Contact

For questions or feedback, please visit gpt-labs.ai