add example models

DocRepairer-XL/results.json

@@ -0,0 +1,19 @@
+{
+  "model_type_symbol": "T",
+  "model": "[DocRepairer-XL](https://huggingface.co/docrepairer-xl)",
+  "average": 0.891,
+  "Type": "seq2seq",
+  "Architecture": "T5-xl",
+  "Weight type": "fine-tuned",
+  "Num tokens": 294500,
+  "Num errors": 482,
+  "CER": 0.043,
+  "WER": 0.071,
+  "CER improvement": 0.034,
+  "WER improvement": 0.051,
+  "Hub License": "mit",
+  "#Params (B)": 3.0,
+  "Hub ❤️": 978,
+  "Available on the hub": true,
+  "Model sha": "d8a9e71"
+}

OCRFixer-Lite/results.json

@@ -0,0 +1,19 @@
+{
+  "model_type_symbol": "T",
+  "model": "[OCRFixer-Lite](https://huggingface.co/ocrfixer-lite)",
+  "average": 0.842,
+  "Type": "seq2seq",
+  "Architecture": "T5-small",
+  "Weight type": "fine-tuned",
+  "Num tokens": 145300,
+  "Num errors": 528,
+  "CER": 0.064,
+  "WER": 0.092,
+  "CER improvement": 0.023,
+  "WER improvement": 0.037,
+  "Hub License": "apache-2.0",
+  "#Params (B)": 0.06,
+  "Hub ❤️": 412,
+  "Available on the hub": true,
+  "Model sha": "5b3f2e1"
+}

generate_results.py

@@ -0,0 +1,69 @@
+import os
+import json
+import re
+
+# Sample data (paste the JSON data from above here)
+models_data = [
+    {
+        "model_type_symbol": "T",
+        "model": "[OCRFixer-Lite](https://huggingface.co/ocrfixer-lite)",
+        "average": 0.842,
+        "Type": "seq2seq",
+        "Architecture": "T5-small",
+        "Weight type": "fine-tuned",
+        "Num tokens": 145300,
+        "Num errors": 528,
+        "CER": 0.064,
+        "WER": 0.092,
+        "CER improvement": 0.023,
+        "WER improvement": 0.037,
+        "Hub License": "apache-2.0",
+        "#Params (B)": 0.06,
+        "Hub ❤️": 412,
+        "Available on the hub": True,
+        "Model sha": "5b3f2e1"
+    },
+    {
+        "model_type_symbol": "T",
+        "model": "[DocRepairer-XL](https://huggingface.co/docrepairer-xl)",
+        "average": 0.891,
+        "Type": "seq2seq",
+        "Architecture": "T5-xl",
+        "Weight type": "fine-tuned",
+        "Num tokens": 294500,
+        "Num errors": 482,
+        "CER": 0.043,
+        "WER": 0.071,
+        "CER improvement": 0.034,
+        "WER improvement": 0.051,
+        "Hub License": "mit",
+        "#Params (B)": 3.0,
+        "Hub ❤️": 978,
+        "Available on the hub": True,
+        "Model sha": "d8a9e71"
+    },
+    # Add the other models similarly...
+]
+
+
+def clean_model_name(model_markdown):
+    """Extract a safe folder name from the markdown model string."""
+    match = re.match(r"\[(.*?)\]", model_markdown)
+    if match:
+        name = match.group(1)
+        # Replace unsafe characters for folder names
+        return re.sub(r'[<>:"/\\|?*]', '_', name)
+    return "unknown_model"
+
+
+for model_info in models_data:
+    folder_name = clean_model_name(model_info["model"])
+    os.makedirs(folder_name, exist_ok=True)
+
+    # Save model_info to results.json in the folder
+    filepath = os.path.join(folder_name, "results.json")
+
+    with open(filepath, "w", encoding="utf-8") as f:
+        json.dump(model_info, f, indent=2, ensure_ascii=False)
+
+print("Folders and results.json files created successfully.")