Upload upload_kaggle.py with huggingface_hub

upload_kaggle.py

@@ -0,0 +1,102 @@
+#!/usr/bin/env python3
+"""
+upload_kaggle.py — Upload the competition dataset to Kaggle
+
+Before running:
+  1. Create your Kaggle API token at https://www.kaggle.com/settings/account
+  2. Set up credentials:
+       mkdir -p ~/.kaggle
+       echo '{"username":"YOUR_USERNAME","key":"YOUR_API_KEY"}' > ~/.kaggle/kaggle.json
+       chmod 600 ~/.kaggle/kaggle.json
+  3. Run: python upload_kaggle.py [--username YOUR_USERNAME]
+
+The dataset directory ./output/clip-pelatnas-p2/ will be uploaded.
+solution.csv is intentionally NOT included in the upload.
+"""
+
+import argparse
+import json
+import subprocess
+import sys
+from pathlib import Path
+
+DATASET_DIR   = Path('./output/clip-pelatnas-p2')
+METADATA_FILE = DATASET_DIR / 'dataset-metadata.json'
+
+
+def check_kaggle_auth():
+    result = subprocess.run(['kaggle', 'whoami'], capture_output=True, text=True)
+    if result.returncode != 0:
+        print("ERROR: Kaggle authentication failed.")
+        print("Set up ~/.kaggle/kaggle.json first.")
+        print("See: https://www.kaggle.com/docs/api")
+        sys.exit(1)
+    username = result.stdout.strip()
+    print(f"Authenticated as: {username}")
+    return username
+
+
+def write_metadata(username: str, dataset_slug: str = 'clip-pelatnas-p2'):
+    metadata = {
+        "title":    "CLIP Pelatnas P2 2026 — ARIA Multimodal Crisis",
+        "id":       f"{username}/{dataset_slug}",
+        "licenses": [{"name": "other"}],
+    }
+    with open(METADATA_FILE, 'w') as f:
+        json.dump(metadata, f, indent=2)
+    print(f"Metadata written: {METADATA_FILE}")
+    return metadata["id"]
+
+
+def upload(dataset_id: str):
+    print(f"\nUploading dataset: {dataset_id}")
+    print(f"Source directory: {DATASET_DIR.absolute()}")
+
+    # Check if dataset already exists
+    check = subprocess.run(
+        ['kaggle', 'datasets', 'list', '--user', dataset_id.split('/')[0], '--search', 'clip-pelatnas-p2'],
+        capture_output=True, text=True
+    )
+
+    if 'clip-pelatnas-p2' in check.stdout:
+        print("Dataset exists — creating a new version...")
+        result = subprocess.run(
+            ['kaggle', 'datasets', 'version', '-p', str(DATASET_DIR),
+             '-m', 'Updated dataset v2', '--dir-mode', 'zip'],
+            check=True
+        )
+    else:
+        print("Creating new dataset...")
+        result = subprocess.run(
+            ['kaggle', 'datasets', 'create', '-p', str(DATASET_DIR), '--dir-mode', 'zip'],
+            check=True
+        )
+
+    print(f"\nUpload complete: https://www.kaggle.com/datasets/{dataset_id}")
+
+
+def main():
+    parser = argparse.ArgumentParser(description='Upload competition dataset to Kaggle')
+    parser.add_argument('--username', help='Kaggle username (optional, read from kaggle.json if not set)')
+    args = parser.parse_args()
+
+    if not DATASET_DIR.exists():
+        print(f"ERROR: Dataset directory not found: {DATASET_DIR}")
+        print("Run prep_dataset.py first.")
+        sys.exit(1)
+
+    if (DATASET_DIR / 'solution.csv').exists():
+        print("WARNING: solution.csv found in dataset directory!")
+        print("Remove it before uploading to prevent leakage.")
+        sys.exit(1)
+
+    username = check_kaggle_auth()
+    if args.username:
+        username = args.username
+
+    dataset_id = write_metadata(username)
+    upload(dataset_id)
+
+
+if __name__ == '__main__':
+    main()