The 2022 TWS AIHPC Technology Forum series, led by experts from TWCC, will guide you through understanding the advantages and importance of TWCC's high-speed AIHPC computing performance for AI deep learning!
As an ML developer or manager, are you often bogged down in the training and combination of various trivial and complicated parameters and algorithms?
The MLflow course in October will effectively solve the above problems for you. Recognizing the difficulty of managing the machine learning process, Taizhi Cloud has integrated MLflow into the OneAI platform, providing traceability and reproducibility in the machine learning process, fully recording all training output data, and allowing for comparison of differences between models through visual charts, effectively increasing the efficiency of machine learning training.
Given the diversity of human language and different usage scenarios, the same message can be expressed in both spoken and written forms, with some text being converted into dates, symbols, etc. Therefore, we will introduce how to use text normalization techniques in NLP to make the messages produced by ASR more readable and improve the accuracy of TTS conversion.
| Date: October 27, 2022 (Thursday)
| Time: 15:00-16:30
| Target audience: Information managers, IT and technology managers, development teams, engineers, business executives, and anyone interested in machine learning applications.
You will learn:
- Manage model training using MLflow through TWCC OneAI service.
- Text normalization for NLP is performed using the Nemo framework via the TWCC CCS service.
Organizer: Taiwan Smart Cloud
Co-organizer: Taiwan Cloud & IoT Industry Association