--- description: Model fine-tuning workflow allowed-tools: Task, Read, Write, Edit, Bash, Grep, Glob argument-hint: --- # Fine-Tuning Workflow Fine-tune for: **$ARGUMENTS** ## Workflow Steps ### Step 1: Data Preparation **Agent:** @fullstack-developer - Collect training data - Format for fine-tuning - Quality validation - Train/val split ### Step 2: Configuration **Agent:** @researcher - Select base model - Configure hyperparameters - Choose PEFT method (LoRA, QLoRA) - Set up training ### Step 3: Training **Agent:** @fullstack-developer - Run fine-tuning - Monitor metrics - Handle checkpoints - Early stopping ### Step 4: Evaluation **Agent:** @tester - Evaluate on test set - Compare to baseline - Check for regressions - Measure improvements ### Step 5: Deployment **Agent:** @fullstack-developer - Merge weights (if LoRA) - Optimize for inference - Deploy model - Monitor performance ## Progress Tracking - [ ] Data prepared - [ ] Configuration complete - [ ] Training finished - [ ] Evaluation passed - [ ] Deployed Execute each step sequentially. Show progress after each step.