Industrial VLM Fine-Tuning

Led the fine-tuning of a 7B-parameter Vision-Language Model using industrial data, successfully surpassing the performance of 72B SOTA models.

VLMFine-TuningIndustrial DataPyTorch

Industrial Vision Language Model (VLM) Fine-Tuning

In the pursuit of optimizing AI for manufacturing, I led a critical project focusing on Vision-Language Models (VLMs) at Pegatron.

Project Highlights

  • Model Fine-Tuning: Directed the fine-tuning process of a 7B-parameter VLM, specifically tailoring it with proprietary industrial data.
  • SOTA Performance: The resulting model achieved exceptional accuracy and efficiency, successfully surpassing the performance of much larger 72B SOTA (State-of-the-Art) models in our specific industrial use cases.