Upload 3 files

extract_odia_ocr_gemini.py

@@ -0,0 +1,282 @@
+"""
+Extract Odia OCR text from benchmark dataset images using Gemini.
+
+This script:
+1) Reads images recursively from benchmark_dataset/images (or a custom directory)
+2) Sends each image to Gemini for OCR
+3) Appends each result row immediately to a CSV file to avoid losing progress
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import os
+from pathlib import Path
+from typing import Any, Iterable
+
+
+DEFAULT_PROMPT = (
+    "You are an OCR assistant for Odia text.\n"
+    "Extract all visible Odia text from this image exactly as written.\n"
+    "Return only the extracted text, without translation or explanation."
+)
+
+SUPPORTED_EXTENSIONS = {".png", ".jpg", ".jpeg", ".webp", ".bmp", ".tiff", ".tif"}
+
+
+def load_dotenv(dotenv_path: Path) -> dict[str, str]:
+    """Parse a simple .env file (KEY=VALUE lines)."""
+    values: dict[str, str] = {}
+    if not dotenv_path.exists():
+        return values
+
+    for raw_line in dotenv_path.read_text(encoding="utf-8").splitlines():
+        line = raw_line.strip()
+        if not line or line.startswith("#") or "=" not in line:
+            continue
+
+        key, value = line.split("=", 1)
+        key = key.strip()
+        value = value.strip().strip("'").strip('"')
+        if key:
+            values[key] = value
+
+    return values
+
+
+def iter_image_paths(images_dir: Path) -> Iterable[Path]:
+    """Yield all supported image files under images_dir recursively."""
+    for path in sorted(images_dir.rglob("*")):
+        if path.is_file() and path.suffix.lower() in SUPPORTED_EXTENSIONS:
+            yield path
+
+
+def call_gemini_ocr(
+    image_path: Path,
+    client: Any,
+    model: str,
+    prompt: str,
+) -> str:
+    """Call Gemini with prompt + image using official google-genai SDK."""
+    try:
+        from PIL import Image
+    except ImportError as exc:
+        raise RuntimeError("Missing dependency: pillow. Install with `pip install pillow`.") from exc
+
+    image = Image.open(image_path).convert("RGB")
+    response = client.models.generate_content(
+        model=model,
+        contents=[prompt, image],
+    )
+    output_text = (response.text or "").strip()
+    if not output_text:
+        raise RuntimeError("Empty OCR output in Gemini response")
+    return output_text
+
+
+def normalize_stored_path(path_str: str, project_root: Path) -> str:
+    """Normalize CSV image_path for stable matching and dedup."""
+    raw = str(path_str).strip()
+    if not raw:
+        return ""
+    p = Path(raw)
+    if p.is_absolute():
+        try:
+            return str(p.resolve().relative_to(project_root))
+        except ValueError:
+            return str(p.resolve())
+    return raw
+
+
+def load_existing_rows_by_path(output_csv: Path, project_root: Path) -> dict[str, dict[str, str]]:
+    """Load CSV rows keyed by normalized image path (latest row wins)."""
+    rows_by_path: dict[str, dict[str, str]] = {}
+    if not output_csv.exists():
+        return rows_by_path
+
+    with output_csv.open("r", encoding="utf-8", newline="") as f:
+        reader = csv.DictReader(f)
+        for row in reader:
+            key = normalize_stored_path(row.get("image_path", ""), project_root)
+            if not key:
+                continue
+            rows_by_path[key] = {
+                "image_path": key,
+                "extracted_odia_text": row.get("extracted_odia_text", "") or "",
+                "status": row.get("status", "") or "",
+                "error": row.get("error", "") or "",
+            }
+    return rows_by_path
+
+
+def image_path_key(image_path: Path, project_root: Path) -> str:
+    """Use project-relative path for CSV storage and deduplication."""
+    resolved = image_path.resolve()
+    try:
+        return str(resolved.relative_to(project_root))
+    except ValueError:
+        return str(resolved)
+
+
+def ensure_output_header(output_csv: Path, append_mode: bool) -> None:
+    """Ensure CSV header exists when creating a new output file."""
+    output_csv.parent.mkdir(parents=True, exist_ok=True)
+    if append_mode and output_csv.exists():
+        return
+    with output_csv.open("w", encoding="utf-8", newline="") as f:
+        writer = csv.writer(f)
+        writer.writerow(["image_path", "extracted_odia_text", "status", "error"])
+
+
+def write_rows(output_csv: Path, rows_by_path: dict[str, dict[str, str]]) -> None:
+    """Rewrite CSV from rows map to keep one row per image path."""
+    output_csv.parent.mkdir(parents=True, exist_ok=True)
+    with output_csv.open("w", encoding="utf-8", newline="") as f:
+        writer = csv.DictWriter(
+            f,
+            fieldnames=["image_path", "extracted_odia_text", "status", "error"],
+        )
+        writer.writeheader()
+        writer.writerows(rows_by_path.values())
+        f.flush()
+
+
+def main() -> None:
+    project_root = Path(__file__).parent.parent
+    dotenv_values = load_dotenv(project_root / ".env")
+
+    default_images_dir = (
+        dotenv_values.get("IMAGE_FOLDER_PATH")
+        or str(project_root / "benchmark_dataset" / "images")
+    )
+    default_output_csv = (
+        dotenv_values.get("OUTPUT_CSV_PATH")
+        or str(project_root / "benchmark_dataset" / "gemini_ocr_output.csv")
+    )
+    default_api_key = dotenv_values.get("GEMINI_API_KEY") or os.getenv(
+        "GEMINI_API_KEY", ""
+    )
+
+    parser = argparse.ArgumentParser(
+        description="Extract Odia OCR text from benchmark images using Gemini"
+    )
+    parser.add_argument(
+        "--model",
+        type=str,
+        default="gemini-3-flash-preview",
+        help="Gemini model name",
+    )
+    parser.add_argument(
+        "--prompt",
+        type=str,
+        default=DEFAULT_PROMPT,
+        help="Prompt used for OCR extraction",
+    )
+    parser.add_argument(
+        "--limit",
+        type=int,
+        default=None,
+        help="Optional max number of images to process",
+    )
+    parser.add_argument(
+        "--no-resume",
+        action="store_true",
+        help="Do not skip already processed image paths in output CSV",
+    )
+    args = parser.parse_args()
+
+    if not default_api_key:
+        raise ValueError(
+            "Gemini API key missing. Set GEMINI_API_KEY in .env or environment."
+        )
+
+    try:
+        from google import genai
+    except ImportError as exc:
+        raise RuntimeError(
+            "Missing dependency: google-genai. Install with `pip install google-genai`."
+        ) from exc
+
+    client = genai.Client(api_key=default_api_key)
+
+    images_dir = Path(default_images_dir).resolve()
+    output_csv = Path(default_output_csv).resolve()
+
+    if not images_dir.exists():
+        raise FileNotFoundError(f"Images directory not found: {images_dir}")
+
+    all_images = list(iter_image_paths(images_dir))
+    if args.limit is not None:
+        all_images = all_images[: max(args.limit, 0)]
+
+    if not all_images:
+        print(f"No images found under: {images_dir}")
+        return
+
+    rows_by_path: dict[str, dict[str, str]] = {}
+    processed_success_paths: set[str] = set()
+    previous_error_rows = 0
+    if not args.no_resume:
+        rows_by_path = load_existing_rows_by_path(output_csv, project_root)
+        processed_success_paths = {
+            p for p, row in rows_by_path.items() if (row.get("status", "").strip().lower() == "ok")
+        }
+        previous_error_rows = sum(
+            1 for row in rows_by_path.values() if row.get("status", "").strip().lower() == "error"
+        )
+    else:
+        ensure_output_header(output_csv, append_mode=False)
+
+    # Deduplicate/normalize existing CSV content on each resume run.
+    if not args.no_resume and output_csv.exists():
+        write_rows(output_csv, rows_by_path)
+
+    existing_keys = set(processed_success_paths)
+
+    to_process = [p for p in all_images if image_path_key(p, project_root) not in existing_keys]
+    total = len(to_process)
+    if total == 0:
+        print("No new images to process. Output CSV is already up to date.")
+        return
+
+    print(f"Found {len(all_images)} images in total")
+    print(f"Already processed successfully: {len(processed_success_paths)}")
+    if previous_error_rows:
+        print(f"Previous error rows available to retry: {previous_error_rows}")
+    print(f"Processing now: {total}")
+    print(f"Writing incremental results to: {output_csv}")
+    for idx, image_path in enumerate(to_process, start=1):
+        image_str = image_path_key(image_path, project_root)
+        status = "ok"
+        extracted_text = ""
+        err = ""
+
+        try:
+            extracted_text = call_gemini_ocr(
+                image_path=image_path,
+                client=client,
+                model=args.model,
+                prompt=args.prompt,
+            )
+        except Exception as exc:  # noqa: BLE001
+            status = "error"
+            err = str(exc)
+
+        # Upsert: keep a single latest row per image path.
+        rows_by_path[image_str] = {
+            "image_path": image_str,
+            "extracted_odia_text": extracted_text,
+            "status": status,
+            "error": err,
+        }
+        write_rows(output_csv, rows_by_path)
+
+        print(f"[{idx}/{total}] {status}: {image_str}")
+
+    print("\nDone.")
+    print(f"Final CSV: {output_csv}")
+
+
+if __name__ == "__main__":
+    main()

inference.ipynb

@@ -0,0 +1,485 @@
+{
+  "cells": [
+    {
+      "cell_type": "code",
+      "execution_count": 3,
+      "id": "384230de",
+      "metadata": {},
+      "outputs": [
+        {
+          "name": "stdout",
+          "output_type": "stream",
+          "text": [
+            "[INFO] Loading dataset...\n",
+            "\n",
+            "[MODEL] OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 0\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 1\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 2\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 3\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 4\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 5\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 6\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 7\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 8\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 9\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 10\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 11\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 12\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 13\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 14\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 15\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 16\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 17\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 18\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 19\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 20\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 21\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 22\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 23\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 24\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 25\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 26\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 27\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 28\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 29\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 30\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 31\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 32 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 33 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 34\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 35\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 36\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 37\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 38\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 39\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 40\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 41\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 42\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 43\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 44\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 45\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 46\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 47\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 48\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 49\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 50\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 51\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 52\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 53\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 54\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 55\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 56\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 57\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 58\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 59\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 60\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 61\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 62\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 63 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 64\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 65\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 66\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 67\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 68\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 69\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 70\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 71\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 72\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 73\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 74\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 75\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 76\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 77\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 78\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 79\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 80\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 81\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 82\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 83\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 84\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 85\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 86\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 87\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 88\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 89\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 90\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 91\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 92\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 93\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 94\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 95\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 96\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 97\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 98\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 99\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 100\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 101\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 102\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 103\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 104\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 105\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 106\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 107\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 108\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 109\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 110\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 111\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 112\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 113\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 114\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 115\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 116\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 117\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 118\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 119\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 120\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 121\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 122\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 123\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 124\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 125\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 126\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 127\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 128\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 129\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 130\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 131\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 132\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 133\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 134\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 135\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 136\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 137\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 138\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 139\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 140\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 141\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 142\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 143\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 144\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 145\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 146\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 147\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 148\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 149 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 150 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 151\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 152\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 153\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 154\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 155\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 156\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 157\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 158\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 159\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 160\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 161\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 162\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 163\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 164\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 165\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 166\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 167\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 168\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 169\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 170 failed: 400 Client Error: Bad Request for url: http://117.18.102.44:34537/v1/chat/completions\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 171\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 172\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 173\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 174\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 175\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 176\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 177\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 178\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 179\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 180\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 181\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 182\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 183\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 184\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 185\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 186\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 187\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 188\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 189\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 190\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 191\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 192\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 193\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 194\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 195\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 196\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 197\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 198\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 199\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 200\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 201\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 202\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 203\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 204\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 205\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 206\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 207\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 208\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 209\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 210 failed: ('Connection aborted.', TimeoutError('timed out'))\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 211 failed: ('Connection aborted.', TimeoutError('timed out'))\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 212 failed: ('Connection aborted.', TimeoutError('timed out'))\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 213 failed: ('Connection aborted.', TimeoutError('timed out'))\n",
+            "✗ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 214 failed: ('Connection aborted.', TimeoutError('timed out'))\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 215\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 216\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 217\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 218\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 219\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 220\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 221\n",
+            "✓ OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3 | sample 222\n"
+          ]
+        }
+      ],
+      "source": [
+        "import os\n",
+        "import re\n",
+        "import time\n",
+        "import base64\n",
+        "import requests\n",
+        "import pandas as pd\n",
+        "\n",
+        "from datasets import load_dataset\n",
+        "from PIL import Image\n",
+        "from io import BytesIO\n",
+        "from rapidfuzz.distance import Levenshtein\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Server Config\n",
+        "# -----------------------------------------\n",
+        "SERVER_IP = \"117.18.102.44\"\n",
+        "PORT = 34537\n",
+        "\n",
+        "MODELS = [\n",
+        "    \"OdiaGenAIOCR/odia-ocr-qwen-finetuned_v3\"\n",
+        "]\n",
+        "\n",
+        "PROMPT = \"Transcribe all the Odia text from this image exactly as it appears. Return only the extracted odia text, nothing else.\"\n",
+        "\n",
+        "OUTPUT_CSV = \"ocr_benchmark_results.csv\"\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Image -> Base64\n",
+        "# -----------------------------------------\n",
+        "def encode_image(image):\n",
+        "    buffer = BytesIO()\n",
+        "    image.save(buffer, format=\"PNG\")\n",
+        "    return base64.b64encode(buffer.getvalue()).decode()\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Metrics\n",
+        "# -----------------------------------------\n",
+        "def compute_cer(gt, pred):\n",
+        "    gt = str(gt)\n",
+        "    pred = str(pred)\n",
+        "    if len(gt) == 0:\n",
+        "        return 0.0\n",
+        "    return Levenshtein.distance(gt, pred) / len(gt)\n",
+        "\n",
+        "\n",
+        "def has_value(value):\n",
+        "    return pd.notna(value) and str(value).strip() != \"\"\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# OCR Server Inference\n",
+        "# -----------------------------------------\n",
+        "def ocr_predict(image, model):\n",
+        "    url = f\"http://{SERVER_IP}:{PORT}/v1/chat/completions\"\n",
+        "    img_base64 = encode_image(image)\n",
+        "\n",
+        "    payload = {\n",
+        "        \"model\": model,\n",
+        "        \"temperature\": 0,\n",
+        "        \"max_tokens\": 1024,\n",
+        "        \"messages\": [\n",
+        "            {\n",
+        "                \"role\": \"user\",\n",
+        "                \"content\": [\n",
+        "                    {\n",
+        "                        \"type\": \"image_url\",\n",
+        "                        \"image_url\": {\n",
+        "                            \"url\": f\"data:image/png;base64,{img_base64}\"\n",
+        "                        },\n",
+        "                    },\n",
+        "                    {\n",
+        "                        \"type\": \"text\",\n",
+        "                        \"text\": PROMPT,\n",
+        "                    },\n",
+        "                ],\n",
+        "            }\n",
+        "        ],\n",
+        "    }\n",
+        "\n",
+        "    resp = requests.post(\n",
+        "        url,\n",
+        "        headers={\n",
+        "            \"Content-Type\": \"application/json\",\n",
+        "            \"Authorization\": \"Bearer EMPTY\",\n",
+        "        },\n",
+        "        json=payload,\n",
+        "        timeout=180,\n",
+        "    )\n",
+        "    resp.raise_for_status()\n",
+        "    return resp.json()[\"choices\"][0][\"message\"][\"content\"].strip()\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Results table helpers\n",
+        "# -----------------------------------------\n",
+        "def model_slug(model_name: str) -> str:\n",
+        "    return re.sub(r\"[^a-zA-Z0-9]+\", \"_\", model_name).strip(\"_\").lower()\n",
+        "\n",
+        "\n",
+        "def model_pred_col(model_name: str) -> str:\n",
+        "    return f\"{model_slug(model_name)}_pred\"\n",
+        "\n",
+        "\n",
+        "def model_cer_col(model_name: str) -> str:\n",
+        "    return f\"{model_slug(model_name)}_cer\"\n",
+        "\n",
+        "\n",
+        "def init_or_load_results(ds):\n",
+        "    base_df = pd.DataFrame(\n",
+        "        {\n",
+        "            \"id\": list(range(len(ds))),\n",
+        "            \"ground_truth\": [row[\"ground_truth\"] for row in ds],\n",
+        "        }\n",
+        "    )\n",
+        "\n",
+        "    if os.path.exists(OUTPUT_CSV):\n",
+        "        existing = pd.read_csv(OUTPUT_CSV)\n",
+        "        if \"id\" not in existing.columns:\n",
+        "            existing = base_df.copy()\n",
+        "        else:\n",
+        "            for col in [\"id\", \"ground_truth\"]:\n",
+        "                if col not in existing.columns:\n",
+        "                    existing[col] = base_df[col]\n",
+        "            existing = existing.sort_values(\"id\", kind=\"stable\")\n",
+        "            existing = existing.drop_duplicates(subset=[\"id\"], keep=\"last\")\n",
+        "            existing = base_df[[\"id\", \"ground_truth\"]].merge(\n",
+        "                existing.drop(columns=[\"ground_truth\"], errors=\"ignore\"),\n",
+        "                on=\"id\",\n",
+        "                how=\"left\",\n",
+        "            )\n",
+        "        return existing\n",
+        "\n",
+        "    return base_df\n",
+        "\n",
+        "\n",
+        "def save_results(df):\n",
+        "    df = df.sort_values(\"id\", kind=\"stable\")\n",
+        "    df.to_csv(OUTPUT_CSV, index=False, encoding=\"utf-8\")\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Benchmark Pipeline\n",
+        "# -----------------------------------------\n",
+        "def run_benchmark(\n",
+        "    hf_repo=\"Iftesha/odia-ocr-benchmark\",\n",
+        "    split=\"train\",\n",
+        "    max_rows=None,\n",
+        "    skip_existing=True,\n",
+        "):\n",
+        "    print(\"[INFO] Loading dataset...\")\n",
+        "    ds = load_dataset(hf_repo, split=split)\n",
+        "\n",
+        "    if max_rows:\n",
+        "        ds = ds.select(range(max_rows))\n",
+        "\n",
+        "    results_df = init_or_load_results(ds)\n",
+        "\n",
+        "    for model in MODELS:\n",
+        "        pred_col = model_pred_col(model)\n",
+        "        cer_col = model_cer_col(model)\n",
+        "\n",
+        "        if pred_col not in results_df.columns:\n",
+        "            results_df[pred_col] = \"\"\n",
+        "        if cer_col not in results_df.columns:\n",
+        "            results_df[cer_col] = \"\"\n",
+        "\n",
+        "        print(f\"\\n[MODEL] {model}\")\n",
+        "        print(f\"[COLUMNS] {pred_col}, {cer_col}\")\n",
+        "\n",
+        "        for i, row in enumerate(ds):\n",
+        "            pred_done = has_value(results_df.at[i, pred_col])\n",
+        "            cer_done = has_value(results_df.at[i, cer_col])\n",
+        "            if skip_existing and pred_done and cer_done:\n",
+        "                print(f\"↷ {model} | sample {i} already exists\")\n",
+        "                continue\n",
+        "\n",
+        "            try:\n",
+        "                image = row[\"image\"]\n",
+        "                if not isinstance(image, Image.Image):\n",
+        "                    image = image.convert(\"RGB\")\n",
+        "\n",
+        "                gt = row[\"ground_truth\"]\n",
+        "                pred = ocr_predict(image, model)\n",
+        "                cer = compute_cer(gt, pred)\n",
+        "\n",
+        "                results_df.at[i, pred_col] = pred\n",
+        "                results_df.at[i, cer_col] = cer\n",
+        "                print(f\"✓ {model} | sample {i}\")\n",
+        "\n",
+        "            except Exception as e:\n",
+        "                print(f\"✗ {model} | sample {i} failed: {e}\")\n",
+        "                # Keep empty value for failed inference so retries can fill it later.\n",
+        "                results_df.at[i, pred_col] = \"\"\n",
+        "                results_df.at[i, cer_col] = \"\"\n",
+        "                time.sleep(1)\n",
+        "\n",
+        "            # Persist after each sample for long-running jobs.\n",
+        "            save_results(results_df)\n",
+        "\n",
+        "    save_results(results_df)\n",
+        "    print(f\"\\n[INFO] Saved comparison CSV to {OUTPUT_CSV}\")\n",
+        "\n",
+        "\n",
+        "# -----------------------------------------\n",
+        "# Run Benchmark\n",
+        "# -----------------------------------------\n",
+        "run_benchmark()"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": null,
+      "id": "3b3e360e",
+      "metadata": {},
+      "outputs": [],
+      "source": []
+    }
+  ],
+  "metadata": {
+    "kernelspec": {
+      "display_name": "Python (vlm-poetry)",
+      "language": "python",
+      "name": "vlm-poetry"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3.14.2"
+    }
+  },
+  "nbformat": 4,
+  "nbformat_minor": 5
+}

upload_to_hf.py

@@ -0,0 +1,257 @@
+"""
+Upload Odia OCR Benchmark Dataset to HuggingFace Hub
+
+Converts local images + metadata.csv to HuggingFace Dataset format and pushes.
+"""
+
+import argparse
+from pathlib import Path
+
+from datasets import Dataset, Features, Value, Image
+from huggingface_hub import HfApi
+import pandas as pd
+
+
+BENCHMARK_DIR = Path(__file__).parent.parent / "benchmark_dataset"
+CSV_PATH = BENCHMARK_DIR / "final_hf.csv"
+
+
+def resolve_image_path(raw_path: str) -> Path:
+    """Resolve image paths from metadata across common path styles."""
+    p = Path(str(raw_path).strip())
+
+    # 1) Already absolute
+    if p.is_absolute():
+        return p
+
+    # 2) Relative to project root: benchmark_dataset/images/...
+    if p.parts and p.parts[0] == "benchmark_dataset":
+        return (BENCHMARK_DIR.parent / p).resolve()
+
+    # 3) Relative to benchmark dir: images/...
+    return (BENCHMARK_DIR / p).resolve()
+
+
+def load_local_dataset() -> Dataset:
+    """Load local images and metadata into a HuggingFace Dataset."""
+    print(f"Loading metadata from {CSV_PATH}...")
+    
+    if not CSV_PATH.exists():
+        raise FileNotFoundError(f"Metadata CSV not found: {CSV_PATH}")
+    
+    df = pd.read_csv(CSV_PATH)
+    print(f"Found {len(df)} samples in metadata")
+    
+    # Normalize image paths, supporting:
+    # - images/...
+    # - benchmark_dataset/images/...
+    # - absolute paths
+    df["image_path"] = df["image_path"].apply(lambda p: str(resolve_image_path(p)))
+    
+    # Verify images exist
+    missing = []
+    for idx, row in df.iterrows():
+        if not Path(row["image_path"]).exists():
+            missing.append(row["image_path"])
+    
+    if missing:
+        print(f"Warning: {len(missing)} images not found:")
+        for p in missing[:5]:
+            print(f"  - {p}")
+        if len(missing) > 5:
+            print(f"  ... and {len(missing) - 5} more")
+        
+        # Filter out missing images
+        df = df[df["image_path"].apply(lambda p: Path(p).exists())]
+        print(f"Continuing with {len(df)} valid samples")
+    
+    if "id" not in df.columns:
+        raise ValueError(
+            f"Required column 'id' not found in {CSV_PATH}. "
+            "Please add an 'id' column before upload."
+        )
+
+    # Create dataset with Image feature
+    features = Features({
+        "id": Value("int64"),
+        "image": Image(),
+        "ground_truth": Value("string"),
+        "category": Value("string"),
+
+    })
+    
+    # Rename image_path to image for HF Dataset
+    data = {
+        "id": df["id"].tolist(),
+        "image": df["image_path"].tolist(),
+        "ground_truth": df["ground_truth"].tolist(),
+        "category": df["category"].tolist(),
+
+    }
+    
+    dataset = Dataset.from_dict(data, features=features)
+    print(f"Created HuggingFace Dataset with {len(dataset)} samples")
+    
+    return dataset
+
+
+def push_to_hub(dataset: Dataset, repo_id: str, private: bool = False):
+    """Push dataset to HuggingFace Hub."""
+    print(f"\nPushing to HuggingFace Hub: {repo_id}")
+    print(f"Private: {private}")
+    
+    dataset.push_to_hub(
+        repo_id,
+        private=private,
+        commit_message="Upload Odia OCR benchmark dataset",
+    )
+    
+    print(f"\nDataset uploaded to: https://huggingface.co/datasets/{repo_id}")
+
+
+def push_dataset_card(repo_id: str, card_content: str):
+    """Upload dataset card as README.md to HuggingFace Hub."""
+    api = HfApi()
+    api.upload_file(
+        path_or_fileobj=card_content.encode("utf-8"),
+        path_in_repo="README.md",
+        repo_id=repo_id,
+        repo_type="dataset",
+        commit_message="Add dataset card README",
+    )
+    print(f"Dataset card uploaded: https://huggingface.co/datasets/{repo_id}/blob/main/README.md")
+
+
+def create_dataset_card(repo_id: str):
+    """Create a dataset card (README.md) for HuggingFace."""
+    card_content = f"""---
+license: cc-by-4.0
+task_categories:
+  - image-to-text
+language:
+  - or
+tags:
+  - ocr
+  - odia
+  - oriya
+  - indic
+  - benchmark
+size_categories:
+  - n<1K
+---
+
+# Odia OCR Benchmark Dataset
+
+## Description
+
+A curated benchmark dataset for evaluating OCR models on Odia (Oriya) text recognition.
+Contains handwritten, printed, scene text, newspaper, books, and digital categories,
+including both short samples and long-text examples for OCR evaluation.
+
+## Dataset Structure
+
+- **id**: Unique identifier for each sample
+- **image**: The input image (PIL Image)
+- **ground_truth**: The correct Odia text transcription
+- **category**: Type of text (handwritten, printed, scene_text, newspaper, books, digital)
+
+
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+dataset = load_dataset("{repo_id}")
+
+# Access a sample
+sample = dataset["train"][0]
+sample_id = sample["id"]
+image = sample["image"]
+text = sample["ground_truth"]
+```
+
+## Categories
+
+| Category      | Description                                           |
+| ------------- | ----------------------------------------------------- |
+| handwritten   | Handwritten Odia text (word/short phrase level)       |
+| printed       | Printed/typed Odia text                               |
+| scene_text    | Text in natural scenes (signboards, posters, etc.)    |
+| newspaper     | Odia newspaper clippings (including long text)        |
+| books         | Scanned Odia book pages (including long text)         |
+| digital       | Screenshots from Odia digital content                 |
+
+## Sources
+
+- `OdiaGenAIOCR/odia-ocr-merged` (handwritten)
+- `darknight054/indic-mozhi-ocr` with config `oriya` (printed)
+- `darknight054/indicstr12-crops` with config `odia` (scene_text)
+- `newspaper`: Odia newspaper scans/clippings
+- `books`: Odia book page images
+- `digital`: odia digital content
+
+## Notes
+
+- Includes long-text samples for paragraph-level OCR evaluation.
+- The `source` field records origin for each sample.
+
+## License
+
+CC-BY-4.0
+"""
+    return card_content
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Upload Odia OCR benchmark dataset to HuggingFace Hub"
+    )
+    parser.add_argument(
+        "--repo",
+        type=str,
+        required=True,
+        help="HuggingFace repo ID (e.g., 'username/odia-ocr-benchmark')",
+    )
+    parser.add_argument(
+        "--private",
+        action="store_true",
+        help="Make the dataset private",
+    )
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="Load and validate dataset without uploading",
+    )
+    
+    args = parser.parse_args()
+    
+    print("=" * 60)
+    print("Upload Odia OCR Benchmark to HuggingFace")
+    print("=" * 60)
+    
+    # Load local dataset
+    dataset = load_local_dataset()
+    
+    # Show sample
+    print("\nSample from dataset:")
+    sample = dataset[0]
+    print(f"  id: {sample['id']}")
+    print(f"  ground_truth: {sample['ground_truth']}")
+    print(f"  category: {sample['category']}")
+
+    
+    if args.dry_run:
+        print("\n[DRY RUN] Dataset validated. Not uploading.")
+        return
+    
+    # Push to hub
+    push_to_hub(dataset, args.repo, private=args.private)
+
+    # Push dataset card
+    card_content = create_dataset_card(args.repo)
+    push_dataset_card(args.repo, card_content)
+
+
+if __name__ == "__main__":
+    main()