Timezone: »

Just Pick a Sign: Optimizing Deep Multitask Models with Gradient Sign Dropout
Zhao Chen · Jiquan Ngiam · Yanping Huang · Thang Luong · Henrik Kretzschmar · Yuning Chai · Dragomir Anguelov

Wed Dec 09 09:00 AM -- 11:00 AM (PST) @ Poster Session 3 #749

The vast majority of deep models use multiple gradient signals, typically corresponding to a sum of multiple loss terms, to update a shared set of trainable weights. However, these multiple updates can impede optimal training by pulling the model in conflicting directions. We present Gradient Sign Dropout (GradDrop), a probabilistic masking procedure which samples gradients at an activation layer based on their level of consistency. GradDrop is implemented as a simple deep layer that can be used in any deep net and synergizes with other gradient balancing approaches. We show that GradDrop outperforms the state-of-the-art multiloss methods within traditional multitask and transfer learning settings, and we discuss how GradDrop reveals links between optimal multiloss training and gradient stochasticity.

Author Information

Zhao Chen (Waymo LLC)
Jiquan Ngiam (Google Brain)
Yanping Huang (Google Brain)
Thang Luong (Google Brain)
Henrik Kretzschmar (Waymo)
Yuning Chai (Waymo)
Dragomir Anguelov (Waymo)

More from the Same Authors