SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2302.04841 · NeurIPS · 2023

Is This Loss Informative? Faster Text-to-Image Customization by Tracking Objective Dynamics

Mipt, Max Ryabinin, Artem Babenko, Anton Voronov, Mikhail Yandex, Khoroshikh

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 2 functions out of this paper's own repositories and ran 2 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
yandex-research/dvar canonical 2 of 2
FunctionStatusWhere it lives
VarEarlyStopper Ran yandex-research/dvar/early_stopping.py
code served (permissive licence) · get_code("47230a89f3ca7c57")
open_image Ran yandex-research/dvar/textual_inversion.py
code served (permissive licence) · get_code("886458c7581cf792")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Text-to-image generation models represent the next step of evolution in image synthesis, offering a natural way to achieve flexible yet fine-grained control over the result. One emerging area of research is the fast adaptation of large text-toimage models to smaller datasets or new visual concepts. However, many efficient methods of adaptation have a long training time, which limits their practical applications, slows down experiments, and spends excessive GPU resources. In this work, we study the training dynamics of popular text-to-image personalization methods (such as Textual Inversion or DreamBooth), aiming to speed them up. We observe that most concepts are learned at early stages and do not improve in quality later, but standard training convergence metrics fail to indicate that. Instead, we propose a simple drop-in early stopping criterion that only requires computing the regular training objective on a fixed set of inputs for all training iterations. Our experiments on Stable Diffusion for 48 different concepts and three personalization methods demonstrate the competitive performance of our approach, which makes adaptation up to 8 times faster with no significant drops in quality.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2302.04841")
get_code_for_paper("2302.04841")
have("2302.04841")

Connect an agent — have() is free.