PromptBoosting: Black-box Text Classification with Ten Forward Passes
Jul 24, 2023
Speakers
Joe O'Connor
Speaker · 0 followers
Jacob Andreas
Speaker · 2 followers
Shiyu Chang
Speaker · 0 followers
Yang Zhang
Speaker · 1 follower
About
We describe PromptBoosting, a query-efficient procedure for building a text classifier from a neural language model (LM) without access to the LM’s parameters, gradients, or hidden representations. This form of “black-box” classifier training has become increasingly important as the cost of training and inference in large-scale LMs has grown. But existing black-box LM classifier learning approaches are themselves computationally inefficient, typically specializing LMs to the target task by searching in a large space of (discrete or continuous) prompts using zeroth-order optimization methods. Instead of directly optimizing in prompt space, PromptBoosting obtains a small pool of prompts via a gradient-free approach and then constructs a large pool of weak learners by pairing these prompts with different elements of the LM’s output distribution. These weak learners are then ensembled using the AdaBoost algorithm. The entire learning process requires only a small number of forward passes and no backward pass. Experiments show that PromptBoosting achieves state-of-the-art performance in multiple black-box few-shot classification tasks, and matches or outperforms full fine-tuning in both few-shot and standard learning paradigms, while training 10x faster than existing black-box methods.We describe PromptBoosting, a query-efficient procedure for building a text classifier from a neural language model (LM) without access to the LM’s parameters, gradients, or hidden representations. This form of “black-box” classifier training has become increasingly important as the cost of training and inference in large-scale LMs has grown. But existing black-box LM classifier learning approaches are themselves computationally inefficient, typically specializing LMs to the target task by searc…
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
WL meet VC
Watch later
Three Dogmas of Reinforcement Learning
Watch later
A Neural PDE Solver with Temporal Stencil Modeling
Watch later
Zhiqing Sun, …
Scaling Class-wise Training Losses for Post-hoc calibration
Watch later
On the Convergence of SARSA with Linear Function Approximation
Watch later
