Standardize script categories and add auto-generated README

amd-arsuresh · amd-arsuresh · commit f130aa0e5555 · 2026-01-16T23:44:02.000+05:30
- Normalized 27 meta.yaml files to use consistent category names
- Removed quotes from all category fields
- Consolidated duplicate categories (ABTF variants)
- Added script/generate_readme.py to auto-generate documentation
- Added update_readme_and_commit.sh for manual updates
- Added git pre-commit hook to auto-update README
- Generated comprehensive script/README.md (308 scripts, 31 categories)
diff --git a/script/README.md b/script/README.md
diff --git a/script/add-custom-nvidia-system/meta.yaml b/script/add-custom-nvidia-system/meta.yaml
@@ -5,7 +5,7 @@ cache: true
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "MLPerf benchmark support"
+category: MLPerf benchmark support
 docker:
   real_run: False
 
diff --git a/script/app-image-classification-onnx-py/meta.yaml b/script/app-image-classification-onnx-py/meta.yaml
@@ -4,7 +4,7 @@ uid: 3d5e908e472b417e
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular AI/ML application pipeline"
+category: Modular AI/ML application pipeline
 
 tags:
 - app
diff --git a/script/app-loadgen-generic-python/meta.yaml b/script/app-loadgen-generic-python/meta.yaml
@@ -5,7 +5,7 @@ uid: d3d949cc361747a6
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf inference benchmark pipeline"
+category: Modular MLPerf inference benchmark pipeline
 
 developers: "[Gaz Iqbal](https://www.linkedin.com/in/gaziqbal), [Arjun Suresh](https://www.linkedin.com/in/arjunsuresh), [Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/app-mlperf-automotive-mlcommons-python/meta.yaml b/script/app-mlperf-automotive-mlcommons-python/meta.yaml
@@ -4,7 +4,7 @@ uid: 621240c5d30a437c
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf inference benchmark pipeline for ABTF model"
+category: Modular MLPerf inference benchmark pipeline
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-automotive/meta.yaml b/script/app-mlperf-automotive/meta.yaml
@@ -4,7 +4,7 @@ uid: f7488ce376484fd2
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf automotive benchmark pipeline for ABTF models"
+category: Modular MLPerf inference benchmark pipeline
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-amd/meta.yaml b/script/app-mlperf-inference-amd/meta.yaml
@@ -6,7 +6,7 @@ cache: false
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf benchmarks"
+category: Modular MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-dummy/meta.yaml b/script/app-mlperf-inference-dummy/meta.yaml
@@ -6,7 +6,7 @@ cache: false
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf benchmarks"
+category: Modular MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-intel/meta.yaml b/script/app-mlperf-inference-intel/meta.yaml
@@ -7,7 +7,7 @@ can_force_cache: true
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf benchmarks"
+category: Modular MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-mlcommons-cpp/meta.yaml b/script/app-mlperf-inference-mlcommons-cpp/meta.yaml
@@ -5,7 +5,7 @@ uid: bf62405e6c7a44bf
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf inference benchmark pipeline"
+category: Modular MLPerf inference benchmark pipeline
 
 developers: "[Thomas Zhu](https://www.linkedin.com/in/hanwen-zhu-483614189), [Arjun Suresh](https://www.linkedin.com/in/arjunsuresh), [Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/app-mlperf-inference-mlcommons-python/meta.yaml b/script/app-mlperf-inference-mlcommons-python/meta.yaml
@@ -5,7 +5,7 @@ uid: ff149e9781fc4b65
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf inference benchmark pipeline"
+category: Modular MLPerf inference benchmark pipeline
 
 developers: "[Arjun Suresh](https://www.linkedin.com/in/arjunsuresh), [Thomas Zhu](https://www.linkedin.com/in/hanwen-zhu-483614189), [Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/app-mlperf-inference-nvidia/meta.yaml b/script/app-mlperf-inference-nvidia/meta.yaml
@@ -7,7 +7,7 @@ can_force_cache: true
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Reproduce MLPerf benchmarks"
+category: Reproduce MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-qualcomm/meta.yaml b/script/app-mlperf-inference-qualcomm/meta.yaml
@@ -7,7 +7,7 @@ can_force_cache: true
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf benchmarks"
+category: Modular MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference-redhat/meta.yaml b/script/app-mlperf-inference-redhat/meta.yaml
@@ -6,7 +6,7 @@ cache: false
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf benchmarks"
+category: Modular MLPerf benchmarks
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/app-mlperf-inference/meta.yaml b/script/app-mlperf-inference/meta.yaml
@@ -5,7 +5,7 @@ uid: d775cac873ee4231
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf inference benchmark pipeline"
+category: Modular MLPerf inference benchmark pipeline
 
 developers: "[Arjun Suresh](https://www.linkedin.com/in/arjunsuresh), [Thomas Zhu](https://www.linkedin.com/in/hanwen-zhu-483614189), [Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/app-mlperf-training-nvidia/meta.yaml b/script/app-mlperf-training-nvidia/meta.yaml
@@ -5,7 +5,7 @@ uid: 1e2e357618cc4674
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf training benchmark pipeline"
+category: Modular MLPerf training benchmark pipeline
 
 # User-friendly tags to find this MLC script
 tags:
diff --git a/script/app-mlperf-training-reference/meta.yaml b/script/app-mlperf-training-reference/meta.yaml
@@ -5,7 +5,7 @@ uid: 0c4b11bdcf494b4f
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular MLPerf training benchmark pipeline"
+category: Modular MLPerf training benchmark pipeline
 
 # User-friendly tags to find this MLC script
 tags:
diff --git a/script/app-stable-diffusion-onnx-py/meta.yaml b/script/app-stable-diffusion-onnx-py/meta.yaml
@@ -4,7 +4,7 @@ uid: 4d33981ac3534b3b
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "Modular AI/ML application pipeline"
+category: Modular AI/ML application pipeline
 
 tags:
 - app
diff --git a/script/build-mlperf-inference-server-nvidia/meta.yaml b/script/build-mlperf-inference-server-nvidia/meta.yaml
@@ -6,7 +6,7 @@ automation_alias: script
 automation_uid: 5b4e0237da074764
 default_version: r3.1
 
-category: "MLPerf benchmark support"
+category: MLPerf benchmark support
 
 
 # User-friendly tags to find this MLC script
diff --git a/script/copy-to-clipboard/meta.yaml b/script/copy-to-clipboard/meta.yaml
@@ -6,7 +6,7 @@ automation_uid: 5b4e0237da074764
 
 cache: false
 
-category: "DevOps automation"
+category: DevOps automation
 
 tags:
 - copy
diff --git a/script/create-patch/meta.yaml b/script/create-patch/meta.yaml
@@ -1,7 +1,7 @@
 uid: 0659dc1f75664c65
 alias: create-patch
 
-category: "DevOps automation"
+category: DevOps automation
 
 automation_alias: script
 automation_uid: 5b4e0237da074764
diff --git a/script/fail/meta.yaml b/script/fail/meta.yaml
@@ -4,7 +4,7 @@ alias: fail
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "DevOps automation"
+category: DevOps automation
 
 cache: false
 
diff --git a/script/generate-mlperf-inference-user-conf/meta.yaml b/script/generate-mlperf-inference-user-conf/meta.yaml
@@ -5,7 +5,7 @@ uid: 3af4475745964b93
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "MLPerf benchmark support"
+category: MLPerf benchmark support
 
 developers: "[Arjun Suresh](https://www.linkedin.com/in/arjunsuresh), [Thomas Zhu](https://www.linkedin.com/in/hanwen-zhu-483614189), [Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/generate-mlperf-tiny-report/meta.yaml b/script/generate-mlperf-tiny-report/meta.yaml
@@ -5,7 +5,7 @@ uid: 709c3f3f9b3e4783
 automation_alias: script
 automation_uid: 5b4e0237da074764
 
-category: "MLPerf benchmark support"
+category: MLPerf benchmark support
 
 developers: "[Grigori Fursin](https://cKnowledge.org/gfursin)"
 
diff --git a/script/generate_readme.py b/script/generate_readme.py
@@ -0,0 +1,169 @@
+#!/usr/bin/env python3
+"""
+Generate README.md for MLCommons scripts directory.
+Scans all meta.yaml files and organizes scripts by category.
+"""
+
+import os
+import yaml
+from pathlib import Path
+from collections import defaultdict
+from datetime import datetime
+
+# Standard category mapping to ensure consistency
+CATEGORY_MAP = {
+    "AI/ML datasets": "AI/ML datasets",
+    "AI/ML frameworks": "AI/ML frameworks", 
+    "AI/ML models": "AI/ML models",
+    "AI/ML optimization": "AI/ML optimization",
+    "Cloud automation": "Cloud automation",
+    "Compiler automation": "Compiler automation",
+    "CUDA automation": "CUDA automation",
+    "Dashboard automation": "Dashboard automation",
+    "Detection or installation of tools and artifacts": "Detection or installation of tools and artifacts",
+    "DevOps automation": "DevOps automation",
+    "Docker automation": "Docker automation",
+    "MLC automation": "MLCommons automation",
+    "MLC Interface": "MLCommons interface",
+    "MLC Script Template": "MLCommons script templates",
+    "MLC Sys Utils": "MLCommons system utilities",
+    "MLC Utils": "MLCommons utilities",
+    "MLPerf benchmark support": "MLPerf benchmark support",
+    "Modular AI/ML application pipeline": "Modular AI/ML application pipeline",
+    "Modular application pipeline": "Modular AI/ML application pipeline",
+    "Modular MLPerf benchmarks": "Modular MLPerf benchmarks",
+    "Modular MLPerf inference benchmark pipeline": "Modular MLPerf inference benchmark pipeline",
+    "Modular MLPerf inference benchmark pipeline for ABTF model": "Modular MLPerf inference benchmark pipeline",
+    "Modular MLPerf automotive benchmark pipeline for ABTF models": "Modular MLPerf inference benchmark pipeline",
+    "Modular MLPerf training benchmark pipeline": "Modular MLPerf training benchmark pipeline",
+    "Platform information": "Platform information",
+    "Python automation": "Python automation",
+    "Remote automation": "Remote automation",
+    "Reproduce MLPerf benchmarks": "Reproduce MLPerf benchmarks",
+    "Reproducibility and artifact evaluation": "Reproducibility and artifact evaluation",
+    "ROCM automation": "ROCm automation",
+    "Tests": "Tests",
+    "TinyML automation": "TinyML automation",
+    "Utils": "Utilities",
+}
+
+def normalize_category(category):
+    """Normalize category name to standard form."""
+    if not category:
+        return "Uncategorized"
+    category = category.strip('"').strip("'").strip()
+    return CATEGORY_MAP.get(category, category)
+
+def read_meta_yaml(script_dir):
+    """Read and parse meta.yaml file."""
+    meta_path = script_dir / "meta.yaml"
+    if not meta_path.exists():
+        return None
+    try:
+        with open(meta_path, 'r', encoding='utf-8') as f:
+            return yaml.safe_load(f)
+    except Exception as e:
+        print(f"Warning: Error reading {meta_path}: {e}")
+        return None
+
+def generate_readme():
+    """Generate README.md from all meta.yaml files."""
+    repo_root = Path(__file__).parent
+    script_dir = repo_root
+    
+    if not script_dir.exists():
+        print(f"Error: script directory not found at {script_dir}")
+        return
+    
+    categories = defaultdict(list)
+    
+    for item in sorted(script_dir.iterdir()):
+        if not item.is_dir() or item.name.startswith('.') or item.name.startswith('_'):
+            continue
+        
+        meta = read_meta_yaml(item)
+        if not meta:
+            continue
+        
+        script_name = item.name
+        display_name = meta.get('name', script_name)
+        category = normalize_category(meta.get('category', 'Uncategorized'))
+        alias = meta.get('alias', script_name)
+        tags = meta.get('tags', [])
+        
+        categories[category].append({
+            'name': script_name,
+            'display_name': display_name,
+            'alias': alias,
+            'tags': tags,
+            'uid': meta.get('uid', ''),
+        })
+    
+    sorted_categories = sorted(categories.items())
+    
+    readme_lines = [
+        "# MLCommons Automation Scripts",
+        "",
+        f"*Last updated: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}*",
+        "",
+        "This directory contains automation scripts for MLPerf benchmarks, AI/ML workflows, and development operations.",
+        "",
+        "## Table of Contents",
+        "",
+    ]
+    
+    for category, _ in sorted_categories:
+        anchor = category.lower().replace(' ', '-').replace('/', '')
+        readme_lines.append(f"- [{category}](#{anchor})")
+    
+    readme_lines.extend(["", "---", ""])
+    
+    for category, scripts in sorted_categories:
+        readme_lines.append(f"## {category}")
+        readme_lines.append("")
+        
+        for script in sorted(scripts, key=lambda x: x['name']):
+            alias_info = f" (alias: `{script['alias']}`)" if script['alias'] != script['name'] else ""
+            readme_lines.append(f"- **[{script['name']}]({script['name']}/)**{alias_info}")
+            readme_lines.append(f"  - {script['display_name']}")
+            if script['tags']:
+                tags_str = ', '.join([f"`{tag}`" for tag in script['tags']])
+                readme_lines.append(f"  - Tags: {tags_str}")
+        
+        readme_lines.append("")
+    
+    readme_lines.extend([
+        "---",
+        "",
+        "## Statistics",
+        "",
+        f"- **Total Scripts**: {sum(len(scripts) for scripts in categories.values())}",
+        f"- **Categories**: {len(categories)}",
+        "",
+        "## Usage",
+        "",
+        "These scripts are part of the MLCommons automation framework. To use them:",
+        "",
+        "```bash",
+        "# Run a script using its alias",
+        "mlc run script <alias> [options]",
+        "",
+        "# Or directly",
+        "cd <script-directory>",
+        "./run.sh [options]",
+        "```",
+        "",
+        "For more information about each script, see the `meta.yaml` file in the script directory.",
+        "",
+    ])
+    
+    readme_path = script_dir / "README.md"
+    with open(readme_path, 'w', encoding='utf-8') as f:
+        f.write('\n'.join(readme_lines))
+    
+    print(f"✓ Generated {readme_path}")
+    print(f"  - {sum(len(scripts) for scripts in categories.values())} scripts")
+    print(f"  - {len(categories)} categories")
+
+if __name__ == "__main__":
+    generate_readme()
diff --git a/script/get-dataset-imagenet-calibration/meta.yaml b/script/get-dataset-imagenet-calibration/meta.yaml
@@ -6,7 +6,7 @@ automation_uid: 5b4e0237da074764
 
 cache: true
 
-category: "AI/ML datasets"
+category: AI/ML datasets
 
 deps:
   - tags: download,file
diff --git a/script/get-dataset-openimages-calibration/meta.yaml b/script/get-dataset-openimages-calibration/meta.yaml
@@ -6,7 +6,7 @@ automation_uid: 5b4e0237da074764
 
 cache: true
 
-category: "AI/ML datasets"
+category: AI/ML datasets
 
 deps:
   - tags: download,file
diff --git a/script/reproduce-mlperf-training-nvidia/meta.yaml b/script/reproduce-mlperf-training-nvidia/meta.yaml
diff --git a/script/runtime-system-infos/meta.yaml b/script/runtime-system-infos/meta.yaml
diff --git a/update_readme_and_commit.sh b/update_readme_and_commit.sh
