compliance-7b-lora

Paralux-Compliance-7B-LoRA

Fine-tuned Mistral-7B-Instruct-v0.3 for financial compliance rule generation.

This model is trained to:

• Convert natural language requirements to structured compliance rules
• Recommend regulatory frameworks based on business context
• Identify compliance gaps in existing rule sets
• Explain regulatory requirements clearly

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Load base model
model = AutoModelForCausalLM.from_pretrained(
    "mistralai/Mistral-7B-Instruct-v0.3",
    torch_dtype=torch.bfloat16,
    device_map="auto"
)

# Load LoRA adapter
model = PeftModel.from_pretrained(model, "paraluxdotai/compliance-7b-lora")
tokenizer = AutoTokenizer.from_pretrained("paraluxdotai/compliance-7b-lora")

# Generate
prompt = """<s>[INST] You are Paralux Compliance Assistant, a specialized AI for financial regulatory compliance.

Convert this requirement to a compliance rule: All financial documents must be retained for 7 years [/INST]"""

inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Details

Intended Use

This model is designed for:

• Financial services compliance automation
• Regulatory rule generation and validation
• Compliance gap analysis
• Policy documentation assistance

Limitations

• Trained specifically on financial compliance data; may not generalize to other domains
• Outputs should be reviewed by compliance professionals before deployment
• Not a substitute for legal advice

License

This LoRA adapter is released under the Apache 2.0 license, matching the base Mistral model.

Citation

@misc{paralux-compliance-7b-lora,
  author = {Paralux},
  title = {Paralux-Compliance-7B-LoRA: Fine-tuned Mistral for Financial Compliance},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/paraluxdotai/compliance-7b-lora}
}