In this episode, we’re joined by Sebastian Ruder, a PhD student studying natural language processing at the National University of Ireland and a Research Scientist at text analysis startup Aylien.
Subscribe: iTunes / Google Play / Spotify / RSS
In our conversation, Sebastian and I discuss recent milestones in neural NLP, including multi-task learning and pretrained language models. We also discuss the use of attention-based models, Tree RNNs and LSTMs, and memory-based networks. Finally, Sebastian walks us through his recent ULMFit paper, short for “Universal Language Model Fine-tuning for Text Classification,” which he co-authored with Jeremy Howard of fast.ai who I interviewed in episode 186.
Thanks to our Sponsor!
I’d like to send a huge thanks to our friends at IBM for their sponsorship of this episode. Are you interested in exploring code patterns leveraging multiple technologies, including ML and AI? Then check out IBM Developer. With more than 100 open source programs, a library of knowledge resources, developer advocates ready to help, and a global community of developers, what in the world will you create? Dive in at ibm.biz/mlaipodcast, and be sure to let them know that TWiML sent you!
Mentioned in the Interview
- Blog Post: A Review of the Recent History of NLP
- Universal Language Model Fine-tuning for Text Classification
- Acoustic Word Embeddings for Low Resource Speech Processing with Herman Kamper
- Paper: A Unified Architecture for Natural Language Processing: Deep Neural Networks with Multitask Learning
- Distributed Representations of Words and Phrases and their Compositionality (Word2vec)
- TWiML Talk #48 – Word2Vec & Friends with Bruno Goncalves
- Neural Turing Machines
- TWiML Talk #186 – The Fastai v1 Deep Learning Framework with Jeremy Howard
- Deep Learning Indaba Series
- TWiML Presents: Series page
- TWiML Events Page
- TWiML Meetup
- TWiML Newsletter
“More On That Later” by Lee Rosevere licensed under CC By 4.0