Deep Double Descent
This video explores a new study on double descent evident in Deep Learning models such as CNNs, ResNets and Transformers. The double descent phenomenon is an interesting intermediate zone where test error increases and then decreases with respect to increasing the model capacity, adding more training data, and training for longer. This is an interesting phenomenon to be aware of when training Deep Learning models!
Paper Link: https://arxiv.org/abs/1912.02292
OpenAI Blog Post: https://openai.com/blog/deep-double-descent/
Thanks for watching! Please Subscribe!
Видео Deep Double Descent канала Connor Shorten
Paper Link: https://arxiv.org/abs/1912.02292
OpenAI Blog Post: https://openai.com/blog/deep-double-descent/
Thanks for watching! Please Subscribe!
Видео Deep Double Descent канала Connor Shorten
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
![Reconciling modern machine learning and the bias-variance trade-off](https://i.ytimg.com/vi/ZAW9EyNo2fw/default.jpg)
![](https://i.ytimg.com/vi/hwnXbnm9Ucw/default.jpg)
![ImageGPT (Generative Pre-training from Pixels)](https://i.ytimg.com/vi/7rFLnQdl22c/default.jpg)
![Most Research in Deep Learning is a Total Waste of Time - Jeremy Howard | AI Podcast Clips](https://i.ytimg.com/vi/Bi7f1JSSlh8/default.jpg)
![SqueezeNet](https://i.ytimg.com/vi/ge_RT5wvHvY/default.jpg)
![Evolution’s great mystery - Michael Corballis](https://i.ytimg.com/vi/lz0lQ58QMzQ/default.jpg)
![Reconciling the double descent curve with older ideas](https://i.ytimg.com/vi/CeghlvXAemc/default.jpg)
![mixup: Beyond Empirical Risk Minimization (Paper Explained)](https://i.ytimg.com/vi/a-VQfQqIMrE/default.jpg)
![Gradient descent, how neural networks learn | Chapter 2, Deep learning](https://i.ytimg.com/vi/IHZwWFHWa-w/default.jpg)
![How Well Can DeepMind's AI Learn Physics? ⚛](https://i.ytimg.com/vi/2Bw5f4vYL98/default.jpg)
![British citizenship by double descent - Romein](https://i.ytimg.com/vi/FuqjmvcmSPo/default.jpg)
![TEDxRedding: Caleen Sisk & Marc Dadigan - Indigenous Knowledge](https://i.ytimg.com/vi/NJGRVW5YWig/default.jpg)
![Datasets for Data-Driven Reinforcement Learning](https://i.ytimg.com/vi/-h1KB8ps11A/default.jpg)
![Elon Musk on Google DeepMind | Artificial Intelligence!](https://i.ytimg.com/vi/MuWWZ91-G6w/default.jpg)
![OpenAI Embeddings API - (Interview Recap and Background)](https://i.ytimg.com/vi/vyf2sqhAm4Y/default.jpg)
![Introduction to Patrilineal Descent w/ Kinship Diagram | Cultural Anthropology](https://i.ytimg.com/vi/Zo8gd4JmsN8/default.jpg)
![Deep Networks Are Kernel Machines (Paper Explained)](https://i.ytimg.com/vi/ahRPdiCop3E/default.jpg)
![OpenAI’s Jukebox AI Writes Amazing New Songs 🎼](https://i.ytimg.com/vi/6oQ0Obi14rM/default.jpg)
![J. Frankle & M. Carbin: The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks](https://i.ytimg.com/vi/s7DqRZVvRiQ/default.jpg)
![The incredible inventions of intuitive AI | Maurice Conti](https://i.ytimg.com/vi/aR5N2Jl8k14/default.jpg)