2022¶

NeMo on the NVIDIA Technical blog in 2022

The following blog posts were published by the NeMo team on the NVIDIA Technical blog in 2022.

August 2022

• Solving Automatic Speech Recognition Deployment Challenges

---

September 2022

• Simplifying Model Development and Building Models at Scale with PyTorch Lightning and NGC

Based on work accepted to Interspeech 2022:

• Improving Japanese Language ASR by Combining Convolutions with Attention Mechanisms
• Changing CTC Rules to Reduce Memory Consumption in Training and Decoding
• Dynamic Scale Weighting Through Multiscale Speaker Diarization
• Text Normalization and Inverse Text Normalization with NVIDIA NeMo

---

NeMo Blog Posts and Announcements

NVIDIA NeMo is a conversational AI toolkit that supports multiple domains such as Automatic Speech Recognition (ASR), Text to Speech generation (TTS), Speaker Recognition (SR), Diarization (SDR), Natural Language Processing (NLP), Neural Machine translation (NMT) and much more. NVIDIA RIVA has long been the toolkit that enables efficient deployment of NeMo models. In recent months, NeMo Megatron supports training and inference on large language models (upto 1 trillion parameters !).

As NeMo becomes capable of more advanced tasks, such as p-tuning / prompt tuning of NeMo Megatron models, domain adaptation of ASR models using Adapter modules, customizable generative TTS models and much more, we introduce this website as a collection of blog posts and announcements for:

• Technical deep dives of NeMo's capabilities
• Presenting state-of-the-art research results
• Announcing new capabilities and domains of research that our team will work on.

Visit NVIDIA NeMo to get started

---