Скачать или смотреть Speech Processing with Deep Learning and JAX/Trax - M. Yusuf Sarıgöz

Speech Processing with Deep Learning and JAX/Trax - M. Yusuf Sarıgöz, co-founder and AI researcher @ AI Labs

With advances in natural language processing, conversational
AI technologies such as personal assistants are getting more and more popular in our everyday life. This requires efficient speech models that can understand and generate human-like speech. In this talk, we will review technologies such as TensorFlow, Jax, Trax, and others that can help boost our research in speech processing. Trax is a new end-to-end deep learning library by Google Brain focusing on clear code and speed, while Jax is "Numpy on steroids." Trax relies on Jax
for hardware acceleration and provides greater productivity for NLP and speech researchers. Finally, we will have a look at the open and free pre-trained German Tacotron2 and Multi-band MelGAN models, their limitations and required future work, and how they were trained.

#machinelearning #Tacotron2 #Trax #germanTTS #melGAN

The timeline:
00:00 Start
00:40 Speaker Introduction
2:00 JAX
5:18 JAX function tracing
6:55 JAX function differentiation
8:22 JAX function jitting
9:20 JAX function vectorization - vmap
12:20 JAX function paralellization - pmap
16:05 Why JAX
19:50 Higher level functions
21:30 Trax introduction
23:15 Why Trax
29:15 German TTS
38:30 Questions