This repository is part of the SimpleMemVLA project, presented in the paper SimpleMemVLA: A Simple but Effective Native-Video Memory for Vision-Language-Action Models.
This repository hosts datasets in LeRobot v3 format for training and evaluating vision-language-action models with long-term memory capabilities. The datasets cover benchmark suites including RMBench, RoboMME, MIKASA-Robo, RoboMemArena, and LIBERO, and are available on Hugging Face under yinchenghust/<benchmark>_lerobot.
huggingface-cli download yinchenghust/libero_lerobot --repo-type dataset --local-dir data/datasets/yinchenghust/libero_lerobot
For full training, evaluation, and benchmark details, see the GitHub repository. This work is released under the MIT License.
This repository is part of the SimpleMemVLA project, presented in the paper SimpleMemVLA: A Simple but Effective Native-Video Memory for Vision-Language-Action Models.
This repository hosts datasets in LeRobot v3 format for training and evaluating vision-language-action models with long-term memory capabilities. The datasets cover benchmark suites including RMBench, RoboMME, MIKASA-Robo, RoboMemArena, and LIBERO, and are available on Hugging Face under yinchenghust/<benchmark>_lerobot.
huggingface-cli download yinchenghust/libero_lerobot --repo-type dataset --local-dir data/datasets/yinchenghust/libero_lerobot
For full training, evaluation, and benchmark details, see the GitHub repository. This work is released under the MIT License.