fairleap-ai/fairleap-v1-clm-sahabatai-8b-adapter

docs: add hugging face model card

feat(training): add the colab finetuning pipeline

feat(eval): add held-out results for all 427 test conversations

feat(eval): add the failure-mode evaluator

feat: add minimal model loading script

feat: add the trained sahabat-ai 8b lora adapter

chore: track model weights and binary assets with git-lfs