RL for Consistency Models: Faster Reward Guided Text-to-Image Generation

Cornell University
RLCM teaser image

Reinforcement Learning for Consistency Models (RLCM). We propose a new framework for finetuning consistency models using RL. On the task of optimizing aesthetic scores of a generated image, comparing to a baseline which uses RL to fine-tune diffusion models (DDPO), RLCM trains (left) and generates images (right) significantly faster, with higher image quality measured under the aesthetic score. Images generated with a batch size of 8 with 8 inference steps.


Reinforcement learning (RL) has improved guided image generation with diffusion models by directly optimizing rewards that capture image quality, aesthetics, and instruction following capabilities. However, the resulting generative policies inherit the same iterative sampling process of diffusion models that causes slow generation. To overcome this limitation, consistency models proposed learning a new class of generative models that directly map noise to data, resulting in a model that can generate an image in as few as one sampling iteration. In this work, to optimize text-to-image generative models for task specific rewards and enable fast training and inference, we propose a framework for fine-tuning consistency models via RL. Our framework, called Reinforcement Learning for Consistency Model (RLCM), frames the iterative inference process of a consistency model as an RL procedure. RLCM improves upon RL fine-tuned diffusion models on text-to-image generation capabilities and trades computation during inference time for sample quality. Experimentally, we show that RLCM can adapt text-to-image consistency models to objectives that are challenging to express with prompting, such as image compressibility, and those derived from human feedback, such as aesthetic qu Comparing to RL finetuned diffusion models, RLCM trains significantly faster, improves the quality of the generation measured under the reward objectives, and speeds up the inference procedure by generating high quality images with as few as two inference steps.

Train Time

Train time
Plots of performance by runtime measured by GPU hours. We report the runtime on four NVIDIA RTX A6000 across three random seeds and plot the mean and standard deviation. We observe that in all tasks RLCM noticeably reduces the training time while achieving comparable or better reward score performance.

Inference Time

Reward time inference
Plots showing the inference performance as a function of time taken to generate. For each task, we evaluated the final checkpoint obtained after training and measured the average score across 100 trajectories at a given time budget on 1 NVIDIA RTX A6000 GPU. We report the mean and std across three seeds for every run. Note that for RLCM, we are able to achieve high scoring trajectories with a smaller inference time budget than DDPO.

Sample Complexity

Sample Complexity
Training curves for RLCM and DDPO by number of reward queries on compressibility, incompressibility, aesthetic, and prompt image alignment. We plot three random seeds for each algorithm and plot the mean and standard deviation across those seeds. RLCM seems to produce either comparable or better reward optimization performance across these tasks.

Qualitative Results

main qualitative results
Representative generations from the pretrained models, DDPO, and RLCM. Across all tasks, we see that RLCM does not compromise the image quality of the base model while being able to transform naturalistic images to be stylized artwork that maximizes an aesthetic score, removes background content to maximize compression, and generate images of animals in fictional scenarios like riding a bike to maximize prompt-alignment.


generalization results
We observe that RLCM is able to generalize to other prompts without substantial decrease in aesthetic quality. The prompts used to test generalization are "bike", "fridge", "waterfall", and "tractor".


        title={RL for Consistency Models: Faster Reward Guided Text-to-Image Generation}, 
        author={Owen Oertell and Jonathan D. Chang and Yiyi Zhang and Kianté Brantley and Wen Sun},