---
title: ChainFlow-VLA-VLM-Feature-Cache
canonical_url: "https://www.modelscope.cn/datasets/AFARI/ChainFlow-VLA-VLM-Feature-Cache"
md_url: "https://www.modelscope.cn/datasets/AFARI/ChainFlow-VLA-VLM-Feature-Cache.md"
repository: AFARI/ChainFlow-VLA-VLM-Feature-Cache
last_updated: 2026-06-03
license: "Apache License 2.0"
storage_size: "964 GB"
downloads: 16547
stars: 0
---

# ChainFlow-VLA-VLM-Feature-Cache

> ChainFlow-VLA-VLM-Feature-Cache - AFARI 在 ModelScope 开源的数据集。ChainFlow-VLA VLM Feature Cache

AFARI/ChainFlow-VLA-VLM-Feature-Cache 是 ModelScope 魔搭社区上的数据集，存储大小 964 GB，采用 Apache License 2.0 许可。

- **Repository**: AFARI/ChainFlow-VLA-VLM-Feature-Cache
- **License**: Apache License 2.0
- **Storage size**: 964 GB
- **Downloads**: 16547
- **Stars**: 0
- **Last updated**: 2026-06-03

Source: https://www.modelscope.cn/datasets/AFARI/ChainFlow-VLA-VLM-Feature-Cache

---

# ChainFlow-VLA VLM Feature Cache

This dataset provides cached VLM hidden features and trajectory targets used by ChainFlow-VLA on NAVSIM.

## Dataset Structure

    navtest/
      navtest.tar.zst
      checksums_navtest.sha256

    trainval/
      trainval_0000.tar.zst
      trainval_0001.tar.zst
      ...
      trainval_0023.tar.zst
      checksums_trainval.sha256
      trainval_logs.txt

The `trainval` split is sharded by NAVSIM log folders. Each shard is an independent `.tar.zst` archive.

After extraction, each token folder contains:

    internvl_feature.gz
    trajectory_target.gz

## Extract

Extract navtest:

    tar --zstd -xf navtest/navtest.tar.zst

Extract all trainval shards:

    for f in trainval/trainval_*.tar.zst; do
        tar --zstd -xf "$f"
    done

## Verify Checksums

Verify navtest:

    cd navtest
    sha256sum -c checksums_navtest.sha256

Verify trainval:

    cd trainval
    sha256sum -c checksums_trainval.sha256

## Notes

- `trainval` contains 24 independent shards.
- `navtest` is packed into a single archive.
- The archives use `zstd` compression. Please install `zstd` before extraction if your system does not support `tar --zstd`.