Transformer ONNX Models

Post by Arjun Kumbakkara / Nabarun Barua Feb 07, 2022 Transformer Optimization

Convert your bulky Transformer models into lightweight high performance ONNX models!

How we converted our ALBERT model trained for text classification to ONNX runtime and how it suddenly increased to 358.3mb from 46.8mb of size( .bin weights file).
...

Arjun Kumbakkara | Author

Senior Enterprise Tech Lead | Gen AI, LLM, DL ,NLP | AIOps & MLOps | AWS | MS AI/ML |Agile Leadership (CSM® Certified)

  • Share this :
← Previous Article Next Article →