Add dataset card for M2Tok data

README.md

@@ -0,0 +1,23 @@
+---
+task_categories:
+- robotics
+---
+
+# M2Tok Data
+
+This repository contains the preprocessed action dataset used in [M2Tok: Multi-head Multi-codebook Discrete Action Tokenization for Vision-Language-Action Models](https://huggingface.co/papers/2609.18259).
+
+Code: https://github.com/cpaaax/M2Tok
+
+## Dataset Preparation
+
+Download `RoboTwin_actions_chunk_8.pkl` from this repository and place it under your data directory. Ensure the `train_data_path` in `action_tokenizer/pretrain_config_epoch200.yaml` is set correctly.
+
+## Sample Usage
+
+Train the action tokenizer using an 8-GPU distributed setup via `torchrun`:
+
+```bash
+cd src/action_tokenizer
+CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 torchrun --nproc_per_node=8 pretrain_vqvae.py config=pretrain_config_epoch200.yaml
+```