Image Classification Model Development

I have a collection of images that needs to be turned into a reliable predictive system—specifically, an image-classification model. The work begins with exploring and cleaning the dataset, then moves through model selection, training, hyper-parameter tuning, and evaluation. I am open to popular frameworks such as TensorFlow, Keras, or PyTorch as long as you can justify the choice and document the pipeline clearly.

Deliverables
• Reproducible code (Python notebooks or scripts)
• Trained model files and clear instructions for inference on new images
• A concise report covering data preparation steps, model architecture, training metrics, and final accuracy/F1-score
• Optional: a lightweight API or CLI for running the classifier locally

Acceptance criteria
• Consistent image classification performance that meets or exceeds the metric targets agreed upon at project kick-off
• Code runs end-to-end on my machine with only the documented dependencies
• Well-commented source so future adjustments are straightforward

If you have previous examples of predictive analysis on visual data, that will speed up the selection process.