Stitching Sub-trajectories with Conditional Diffusion Model for Goal-Conditioned Offline RL

Authors: Sungyoon Kim, Yunseon Choi, Daiki E. Matsunaga, Kee-Eung Kim

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We report state-of-the-art performance in the standard benchmark set of GCRL tasks, and demonstrate the capability to successfully stitch the segments of suboptimal trajectories in the offline data to generate highquality plans. In this section, we demonstrate the effectiveness of the proposed SSD approach in two different GCRL domains: Maze2D and Fetch.
Researcher Affiliation Academia Kim Jaechul Graduate School of AI, KAIST {sykim, yschoi, dematsunaga}@ai.kaist.ac.kr, kekim@kaist.ac.kr
Pseudocode Yes Algorithm 1: SSD (Training)
Open Source Code Yes Our code is available publicly at https: //github.com/rlatjddbs/SSD
Open Datasets Yes We utilize the D4RL dataset (Fu et al. 2020), which is generated by a hand-designed PID controller as a planner, which produces a sequence of waypoints.
Dataset Splits No The paper mentions total dataset sizes but does not specify exact training, validation, and test split percentages or sample counts for the experiments.
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, memory) used for running experiments were mentioned in the paper.
Software Dependencies No The paper does not provide specific software dependencies with version numbers needed to replicate the experiment.
Experiment Setup No The paper describes the overall training procedure but does not provide specific hyperparameter values (e.g., learning rate, batch size, number of epochs) or detailed training configurations in the main text.