3 sessions on huggingface datasets with a coding agent: the agent failed at least once in 2, and 1 finished without a human stepping in.
Takeaway: Next time I'd front-load the dataset ID and scope questions in the first prompt so the agent doesn't have to burn a dozen turns guessing at what I meant by 'seed_dict' vs full retrain.
Takeaway: Next time I'd have it check for a working Python runtime and gradio version up front, since that cost several minutes of failed docs searches and PATH issues before it could even inspect the API.
Takeaway: Next time I'd tell the agent upfront to check the model card on Hugging Face for the exact transformers version needed, since half the debugging loop was chasing version mismatches (MiniCPM-V-4.6 model_type not registered, then peft/transformers pin conflicts) that a single doc check would have caught.