JurisQwen
Qwen2.5-7B fine-tuned on Indian legal datasets with LoRA and 4-bit quantization, deployed on Modal.
Qwen2.5-7B fine-tuned on Indian legal datasets using LoRA + PEFT + Unsloth. Deployed with 4-bit quantization and Flash Attention 2 on Modal for legal document analysis and Q&A.