--- base_model: Qwen/Qwen3.5-2B base_model_relation: finetune library_name: transformers tags: - clinlab - clinverse - clinical-laboratory - merged --- # ClinLab-2B ClinLab-2B is a Qwen3.5-2B model fine-tuned for clinical laboratory simulation in ClinVerse using the MIMIC-IV-Ext training pipeline. This repository contains the full merged model weights, tokenizer, chat template, and model configuration. The final LoRA adapter has already been merged into the base model; a separate adapter is not required. ## Files and provenance - Base model: [Qwen/Qwen3.5-2B](https://huggingface.co/Qwen/Qwen3.5-2B). - Export: merged MIMIC-IV-Ext checkpoint. - Weight format: Safetensors. - Stored precision: bfloat16. - Training records, patient records, and optimizer checkpoints are not included. Use the tokenizer chat template and the laboratory-provider prompt format from ClinVerse for the intended task. This checkpoint has not been evaluated as a general-purpose medical assistant.