manifest.yaml

onescience_manifest_version: "0.1"
resource_type: dataset
resource:
  id: OneScience/cfd_benchmark
  name: cfd_benchmark
  domain: cfd
  domain_tags:
    - cfd
    - neural_pde_solver
    - benchmark
  task: cfd_operator_learning_dataset
  task_tags:
    - flow_field_prediction
    - pde_operator_learning
    - data_reading_validation
  modalities:
    - gridded_field
    - unstructured_mesh
  input_formats:
    - npy
    - mat
  output_formats: []
  summary: CFD_Benchmark 数据集的 ModelScope 标准数据包，本次包含 airfoil、darcy、elasticity、ns、pipe、plas 六个子集。

platform_resource:
  primary:
    platform: modelscope
    repo_id: OneScience/cfd_benchmark
    repo_type: dataset
    url: https://modelscope.cn/datasets/OneScience/cfd_benchmark
    revision: main
    readme_path: README.md
    manifest_path: manifest.yaml
  mirrors: []
  access:
    visibility: public
    license: null

website_integration:
  enabled: true
  click_target:
    platform: modelscope
    resource_url: https://modelscope.cn/datasets/OneScience/cfd_benchmark
  llm_handoff:
    readme_required: true
    manifest_required: true
    download_readme_first: true
    resolve_related_models: true
    default_run_goal: minimal_validation

runtime:
  enabled: true
  onescience_domain: cfd
  min_onescience_version: null
  supported_execution:
    - local
    - local_slurm
    - remote_slurm

onescience:
  repo: https://gitee.com/onescience-ai/onescience
  official_links:
    gitee:
      doc: https://gitee.com/onescience-ai/onescience-doc
      onescience: https://gitee.com/onescience-ai/onescience
      skills: https://gitee.com/onescience-ai/oneskills
    github:
      doc: https://github.com/onescience-ai/OneScience-doc
      onescience: https://github.com/onescience-ai/OneScience
      skills: https://github.com/onescience-ai/oneskills
  install:
    required_by_default: false
    command: source ~/env.sh && bash install.sh cfd
  source_paths: []
  compatibility:
    examples_path: examples/cfd/CFD_Benchmark
    status: examples_compatible

runtime_package:
  kind: standard_runtime_package
  package_root: .
  standard_layout:
    workdir: .
    config_dir: metadata
    checkpoint_dir: null
    data_dir: data
    output_dir: reports
  apply_policy:
    mode: direct_use
    target: session_workdir
    overwrite: true
    protect_installed_onescience: true
  entry_files:
    - metadata/file_manifest.csv
    - metadata/schema.yaml
    - scripts/validate_dataset.py
  entrypoints:
    preflight: scripts/validate_dataset.py
    inference: null
    train: null
    finetune: null
    evaluate: null
    visualize: null
    deploy: null

files:
  model_files: []
  dataset_files:
    - id: cfd_benchmark_data
      path: data/
      role: train_data
      description_zh: 本次整理的 CFD_Benchmark 六个子集数据目录，包含 airfoil、darcy、elasticity、ns、pipe、plas。
      required: true
      required_for:
        - train
        - evaluate
        - preflight
      format: npy_mat
      source:
        platform: modelscope
        repo_id: OneScience/cfd_benchmark
        repo_type: dataset
        path: data/
        revision: main
        download_method: command_ref
        command_ref: commands.download.download_dataset
      local_path: session_workdir/cfd_cfd_benchmark_data/data/
      size: "约 8.2G"
      checksum: metadata/file_manifest.csv
      large_file: true
    - id: file_manifest
      path: metadata/file_manifest.csv
      role: checksum_manifest
      description_zh: 整理后每个数据文件的相对路径、字节数和 SHA256。
      required: true
      required_for:
        - preflight
      format: csv
      source:
        platform: modelscope
        repo_id: OneScience/cfd_benchmark
        repo_type: dataset
        path: metadata/file_manifest.csv
        revision: main
        download_method: command_ref
        command_ref: commands.download.download_dataset
      local_path: session_workdir/cfd_cfd_benchmark_data/metadata/file_manifest.csv
      size: null
      checksum: null
      large_file: false
  config_files:
    - id: schema
      path: metadata/schema.yaml
      role: schema
      description_zh: 记录 npy shape/dtype 和 MATLAB 文件变量结构。
      required: true
      required_for:
        - preflight
        - train
        - evaluate
      format: yaml
      source:
        platform: modelscope
        repo_id: OneScience/cfd_benchmark
        repo_type: dataset
        path: metadata/schema.yaml
        revision: main
        download_method: command_ref
        command_ref: commands.download.download_dataset
      local_path: session_workdir/cfd_cfd_benchmark_data/metadata/schema.yaml
      checksum: null
      large_file: false
  sample_files: []

dataset:
  format:
    - npy
    - matlab_v5_mat
  sample_unit: 不同子集的一个样本定义不同，airfoil/pipe 为一个几何流场样本，darcy/ns/plas 为一个 PDE 系数或时间序列样本，elasticity 为一个随机单元结构样本。
  schema:
    path: metadata/schema.yaml
  checksum_manifest:
    path: metadata/file_manifest.csv
  splits:
    train: 由 CFD_Benchmark 运行脚本通过 ntrain 或 train_ratio 切分。
    validation: null
    test: 由 CFD_Benchmark 运行脚本通过 ntest 或 train_ratio 切分。

relations:
  required_datasets: []
  optional_datasets: []
  compatible_models:
    - id: OneScience/CFD_Benchmark
      role: train_data
      required_for:
        - train
        - evaluate
        - preflight
      resource_ref:
        platform: modelscope
        repo_id: OneScience/CFD_Benchmark
        repo_type: model
        url: https://modelscope.cn/models/OneScience/CFD_Benchmark
        revision: main
        readme_path: README.md
        manifest_path: manifest.yaml
      expected_local_path: ../cfd_cfd_benchmark_data/data
      adapter: scripts/standard_train_airfoil_transolver.sh

run_matrix:
  scenarios:
    - name: minimal_validation
      capability: preflight
      default: true
      required_datasets:
        - id: OneScience/cfd_benchmark
          role: train_data
          local_path: data/
      required_model_files: []
      command_refs:
        - commands.preflight.validate_dataset
      expected_outputs:
        - reports/dataset_validation.json
      prerequisites:
        - 已下载数据集 OneScience/cfd_benchmark。
    - name: cfd_benchmark_model_train
      capability: train
      required_datasets:
        - id: OneScience/cfd_benchmark
          role: train_data
          local_path: data/
      required_model_files:
        - OneScience/CFD_Benchmark runtime/
      command_refs:
        - commands.preflight.validate_dataset
      expected_outputs:
        - reports/dataset_validation.json
      prerequisites:
        - 下载并进入模型仓库 OneScience/CFD_Benchmark。

capabilities:
  inference: false
  train: false
  finetune: false
  evaluate: false
  visualize: false
  deploy: false
  data_reading_validation: true

commands:
  download:
    - name: download_dataset
      description: 下载 CFD_Benchmark 数据集标准包。
      cwd: .
      command: modelscope download --dataset OneScience/cfd_benchmark --local_dir ./cfd_cfd_benchmark_data
      required_files: []
      required_env:
        - MODELSCOPE_TOKEN
      expected_outputs:
        - path: cfd_cfd_benchmark_data/manifest.yaml
          type: file
        - path: cfd_cfd_benchmark_data/data/
          type: directory
      success_criteria:
        - exit_code == 0
  preflight:
    - name: validate_dataset
      description: 校验 14 个数据文件的存在性、大小、SHA256、npy shape/dtype 和 mat 变量结构。
      cwd: .
      command: source ~/env.sh && python scripts/validate_dataset.py --data-root data --file-manifest metadata/file_manifest.csv
      required_files:
        - scripts/validate_dataset.py
        - metadata/file_manifest.csv
        - data/
      required_env:
        - python
        - numpy
        - scipy
      expected_outputs:
        - path: reports/dataset_validation.json
          type: file
      success_criteria:
        - exit_code == 0
  prepare: []
  inference: []
  train: []
  finetune: []
  evaluate: []
  visualize: []
  deploy: []

expected_outputs:
  - path: reports/dataset_validation.json
    type: json
    description_zh: 数据文件完整性和可读性验证报告。
  - path: metadata/file_manifest.csv
    type: csv
    description_zh: 文件名、大小和 SHA256 校验清单。
  - path: metadata/schema.yaml
    type: yaml
    description_zh: 数据结构说明。

diagnostics:
  - id: missing_file
    symptom: validate_dataset.py 报告 status=missing。
    cause: 数据下载不完整或放置目录错误。
    fix: 重新执行 modelscope download --dataset OneScience/cfd_benchmark --local_dir ./cfd_cfd_benchmark_data。
  - id: sha256_mismatch
    symptom: validate_dataset.py 报告 status=sha256_mismatch。
    cause: 文件损坏、复制中断或上传下载过程中内容变化。
    fix: 删除对应文件后重新下载数据集，或用 metadata/source_file_manifest.csv 与源目录重新比对。
  - id: scipy_missing
    symptom: MATLAB 文件只完成 header_only 检查。
    cause: 当前 Python 环境缺少 scipy。
    fix: source ~/env.sh 后重跑验证；当前标准化过程中已用 scipy 生成 metadata/schema.yaml。

domain_extension:
  cfd:
    benchmark_name: CFD_Benchmark
    included_subsets:
      airfoil:
        files:
          - NACA_Cylinder_X.npy
          - NACA_Cylinder_Y.npy
          - NACA_Cylinder_Q.npy
      darcy:
        files:
          - piececonst_r421_N1024_smooth1.mat
          - piececonst_r421_N1024_smooth2.mat
      elasticity:
        files:
          - Meshes/Random_UnitCell_XY_10.npy
          - Meshes/Random_UnitCell_rr_10.npy
          - Meshes/Random_UnitCell_sigma_10.npy
          - Meshes/Random_UnitCell_theta_10.npy
      ns:
        files:
          - NavierStokes_V1e-5_N1200_T20.mat
      pipe:
        files:
          - Pipe_X.npy
          - Pipe_Y.npy
          - Pipe_Q.npy
      plas:
        files:
          - plas_N987_T20.mat
    skipped_source_subsets:
      - car
      - diffusion_sorption