add register_for_auto_class("AutoTokenizer")

app.py

@@ -341,12 +341,20 @@ def evaluate_legal_moves(
     """Evaluate a model's legal move generation."""
     try:
         import sys
+        import io
+        from contextlib import redirect_stdout
+        
         sys.path.insert(0, str(Path(__file__).parent))
         
         from src.evaluate import ChessEvaluator, load_model_from_hub
         
         progress(0, desc="Loading model...")
-        model, tokenizer = load_model_from_hub(model_id)
+        
+        # Capture tokenizer debug info
+        debug_output = io.StringIO()
+        with redirect_stdout(debug_output):
+            model, tokenizer = load_model_from_hub(model_id, verbose=True)
+        tokenizer_info = debug_output.getvalue()
         
         progress(0.1, desc="Setting up evaluator...")
         evaluator = ChessEvaluator(
@@ -409,6 +417,9 @@ which adds the required metadata to the README.md file.
         
         progress(1.0, desc="Done!")
         
+        # Format tokenizer info for display
+        tokenizer_debug = tokenizer_info.strip().replace("   ", "- ")
+        
         return f"""
 ## Legal Move Evaluation for {model_id.split('/')[-1]}
 
@@ -419,6 +430,11 @@ which adds the required metadata to the README.md file.
 | **Legal (with retries)** | {results['legal_first_try'] + results['legal_with_retry']} ({results['legal_rate_with_retry']*100:.1f}%) |
 | **Always Illegal** | {results['illegal_all_retries']} ({results['illegal_rate']*100:.1f}%) |
 
+### Tokenizer Info
+```
+{tokenizer_debug}
+```
+
 ### Leaderboard Update
 {update_message}
 

src/evaluate.py

@@ -750,13 +750,14 @@ class ChessEvaluator:
         return results
 
 
-def load_model_from_hub(model_id: str, device: str = "auto"):
+def load_model_from_hub(model_id: str, device: str = "auto", verbose: bool = True):
     """
     Load a model from the Hugging Face Hub.
     
     Args:
         model_id: Model ID on Hugging Face Hub.
         device: Device to load the model on.
+        verbose: Whether to print debug info about loaded tokenizer.
     
     Returns:
         Tuple of (model, tokenizer).
@@ -769,10 +770,15 @@ def load_model_from_hub(model_id: str, device: str = "auto"):
     
     # Try AutoTokenizer with trust_remote_code first to load custom tokenizer.py from Hub
     # Fall back to local ChessTokenizer if the model doesn't have a custom tokenizer
+    tokenizer_source = None
     try:
         tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
-    except Exception:
+        tokenizer_source = "AutoTokenizer (from Hub with trust_remote_code=True)"
+    except Exception as e:
+        if verbose:
+            print(f"   AutoTokenizer failed: {e}")
         tokenizer = ChessTokenizer.from_pretrained(model_id)
+        tokenizer_source = "ChessTokenizer (local class, vocab from Hub)"
     
     model = AutoModelForCausalLM.from_pretrained(
         model_id,
@@ -780,6 +786,15 @@ def load_model_from_hub(model_id: str, device: str = "auto"):
         device_map=device,
     )
     
+    # Print debug info
+    if verbose:
+        print(f"   Tokenizer loaded via: {tokenizer_source}")
+        print(f"   Tokenizer class: {type(tokenizer).__name__}")
+        print(f"   Tokenizer vocab size: {tokenizer.vocab_size}")
+        # Check if tokenizer has custom attributes that might differ
+        if hasattr(tokenizer, '_vocab'):
+            print(f"   Tokenizer has _vocab attribute: yes ({len(tokenizer._vocab)} entries)")
+    
     return model, tokenizer
 
 

submit.py

@@ -78,6 +78,10 @@ def main():
     with tempfile.TemporaryDirectory() as tmp_dir:
         tmp_path = Path(tmp_dir)
 
+        # Register tokenizer for AutoTokenizer so it can be loaded with trust_remote_code=True
+        # This adds the 'auto_map' field to tokenizer_config.json
+        tokenizer.register_for_auto_class("AutoTokenizer")
+        
         # Save model and tokenizer
         model.save_pretrained(tmp_path)
         tokenizer.save_pretrained(tmp_path)