Hasnat contributed to the ml-explore/mlx-swift-examples repository by implementing DeepSeek model configuration within the LLMModelFactory, enabling more flexible and scalable large language model experiments. Using Swift and leveraging skills in dependency management and machine learning, Hasnat updated the swift-transformers and jinja dependencies to support seamless DeepSeek integration. This work established a foundation for expanded model configurability, reducing future integration risks and streamlining onboarding for researchers. Although the contribution focused on feature development rather than bug fixes, the depth of the changes improved the project’s readiness for future releases and accelerated experimentation with configurable LLM deployments in Swift environments.