Chess Challenge submission by alexandreduplessis

README.md

@@ -0,0 +1,26 @@
+---
+library_name: transformers
+tags:
+- chess
+- llm-course
+- chess-challenge
+license: mit
+---
+
+# chess_basic_tokenizer
+
+Chess model submitted to the LLM Course Chess Challenge.
+
+## Submission Info
+
+- **Submitted by**: [alexandreduplessis](https://huggingface.co/alexandreduplessis)
+- **Parameters**: 714,112
+- **Organization**: LLM-course
+
+## Model Details
+
+- **Architecture**: Chess Transformer (GPT-style)
+- **Vocab size**: 153
+- **Embedding dim**: 128
+- **Layers**: 4
+- **Heads**: 4

config.json

@@ -0,0 +1,20 @@
+{
+  "architectures": [
+    "ChessForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "dropout": 0.1,
+  "eos_token_id": 2,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "chess_transformer",
+  "n_ctx": 256,
+  "n_embd": 128,
+  "n_head": 4,
+  "n_inner": 384,
+  "n_layer": 4,
+  "pad_token_id": 0,
+  "tie_weights": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.55.0",
+  "vocab_size": 153
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca2797084e913631bccb1517918a225573b12c3eb0c789a4c989be718051f1d2
+size 2860848

special_tokens_map.json

@@ -0,0 +1,6 @@
+{
+  "bos_token": "[BOS]",
+  "eos_token": "[EOS]",
+  "pad_token": "[PAD]",
+  "unk_token": "[UNK]"
+}

tokenizer.py

@@ -0,0 +1,356 @@
+"""
+Custom Chess Tokenizer for the Chess Challenge.
+
+This tokenizer treats each move as a single token using the extended UCI notation
+from the Lichess dataset (e.g., WPe2e4, BNg8f6).
+
+The dataset format uses:
+- W/B prefix for White/Black
+- Piece letter: P=Pawn, N=Knight, B=Bishop, R=Rook, Q=Queen, K=King
+- Source and destination squares (e.g., e2e4)
+- Special suffixes: (x)=capture, (+)=check, (+*)=checkmate, (o)/(O)=castling
+"""
+
+from __future__ import annotations
+
+import json
+import os
+from pathlib import Path
+from typing import Dict, List, Optional
+
+from transformers import PreTrainedTokenizer
+
+import re
+
+
+
+class ChessTokenizer(PreTrainedTokenizer):
+    """
+    A custom tokenizer for chess moves using extended UCI notation.
+    
+    This tokenizer maps each possible chess move to a unique token ID.
+    The vocabulary is built from the training dataset to ensure all moves
+    encountered during training have a corresponding token.
+    
+    Example:
+        >>> tokenizer = ChessTokenizer()
+        >>> tokenizer.encode("WPe2e4 BPe7e5")
+        [1, 42, 87, 2]  # [BOS, e2e4, e7e5, EOS]
+    """
+    
+    model_input_names = ["input_ids", "attention_mask"]
+    vocab_files_names = {"vocab_file": "vocab.json"}
+    
+    # Special tokens
+    PAD_TOKEN = "[PAD]"
+    BOS_TOKEN = "[BOS]"
+    EOS_TOKEN = "[EOS]"
+    UNK_TOKEN = "[UNK]"
+    
+    _MOVE_RE = re.compile(
+        r'^(?P<color>[WB])(?P<piece>[PNBRQK])(?P<from>[a-h][1-8])(?P<to>[a-h][1-8])(?P<rest>.*)$'
+    )
+
+    _SUFFIX_MAP = {
+        "(x)": "cap",
+        "(+)": "check",
+        "(+*)": "mate",
+        "(o)": "castle_k",
+        "(O)": "castle_q",
+    }
+
+    _PROMO_RE = re.compile(r'=?([QRBNqrbn])')  # accepts "=Q" or "q" style
+
+
+    def __init__(
+        self,
+        vocab_file: Optional[str] = None,
+        vocab: Optional[Dict[str, int]] = None,
+        **kwargs,
+    ):
+        """
+        Initialize the chess tokenizer.
+        
+        Args:
+            vocab_file: Path to a JSON file containing the vocabulary mapping.
+            vocab: Dictionary mapping tokens to IDs (alternative to vocab_file).
+            **kwargs: Additional arguments passed to PreTrainedTokenizer.
+        """
+        # Initialize special tokens
+        self._pad_token = self.PAD_TOKEN
+        self._bos_token = self.BOS_TOKEN
+        self._eos_token = self.EOS_TOKEN
+        self._unk_token = self.UNK_TOKEN
+
+        # Remove any duplicate special-token entries passed through kwargs
+        # to avoid "multiple values for keyword" errors when loading from disk.
+        kwargs.pop("pad_token", None)
+        kwargs.pop("bos_token", None)
+        kwargs.pop("eos_token", None)
+        kwargs.pop("unk_token", None)
+        
+        # Load or create vocabulary
+        if vocab is not None:
+            self._vocab = vocab
+        elif vocab_file is not None and os.path.exists(vocab_file):
+            with open(vocab_file, "r", encoding="utf-8") as f:
+                self._vocab = json.load(f)
+        else:
+            # Create a minimal vocabulary with just special tokens
+            # The full vocabulary should be built from the dataset
+            self._vocab = self._create_default_vocab()
+        
+        # Create reverse mapping
+        self._ids_to_tokens = {v: k for k, v in self._vocab.items()}
+        
+        # Call parent init AFTER setting up vocab
+        super().__init__(
+            pad_token=self._pad_token,
+            bos_token=self._bos_token,
+            eos_token=self._eos_token,
+            unk_token=self._unk_token,
+            **kwargs,
+        )
+    
+    def _decompose_move(self, tok: str) -> List[str]:
+        """
+        Convert e.g. 'WPe2e4(x)' -> ['WP', 'e2_f', 'e4_t', 'cap']
+        """
+        m = self._MOVE_RE.match(tok)
+        if not m:
+            return [self.UNK_TOKEN]
+
+        color = m.group("color")
+        piece = m.group("piece")
+        from_sq = m.group("from")
+        to_sq = m.group("to")
+        rest = m.group("rest") or ""
+
+        out = [f"{color}{piece}", f"{from_sq}_f", f"{to_sq}_t"]
+
+        for raw, mapped in self._SUFFIX_MAP.items():
+            if raw in rest:
+                out.append(mapped)
+
+        # Promotion (rare depending on dataset formatting, but safe)
+        pm = self._PROMO_RE.search(rest)
+        if pm:
+            p = pm.group(1).lower()
+            if p in ("q", "r", "b", "n"):
+                out.append(f"promo_{p}")
+
+        return out
+    def _create_default_vocab(self) -> Dict[str, int]:
+        """
+        Create a minimal default vocabulary with just special tokens.
+        
+        For the full vocabulary, use `build_vocab_from_dataset()`.
+        This minimal vocab is just a placeholder - you should build from data.
+        """
+        special_tokens = [self.PAD_TOKEN, self.BOS_TOKEN, self.EOS_TOKEN, self.UNK_TOKEN]
+        vocab = {token: idx for idx, token in enumerate(special_tokens)}
+        return vocab
+    
+    @classmethod
+    def build_vocab_from_iterator(
+        cls,
+        iterator,
+        min_frequency: int = 1,
+    ) -> "ChessTokenizer":
+        """
+        Build a tokenizer vocabulary from an iterator of game strings.
+        
+        Args:
+            iterator: An iterator yielding game strings (space-separated moves).
+            min_frequency: Minimum frequency for a token to be included.
+        
+        Returns:
+            A ChessTokenizer with the built vocabulary.
+        """
+        from collections import Counter
+        
+        token_counts = Counter()
+        
+        for game in iterator:
+            moves = game.strip().split()
+            token_counts.update(moves)
+        
+        # Filter by frequency
+        tokens = [
+            token for token, count in token_counts.items()
+            if count >= min_frequency
+        ]
+        
+        # Sort for reproducibility
+        tokens = sorted(tokens)
+        
+        # Build vocabulary
+        special_tokens = [cls.PAD_TOKEN, cls.BOS_TOKEN, cls.EOS_TOKEN, cls.UNK_TOKEN]
+        vocab = {token: idx for idx, token in enumerate(special_tokens + tokens)}
+        
+        return cls(vocab=vocab)
+    
+    @classmethod
+    def build_vocab_from_dataset(
+        cls,
+        dataset_name: str = "dlouapre/lichess_2025-01_1M",
+        split: str = "train",
+        column: str = "text",
+        min_frequency: int = 500,
+        max_samples: Optional[int] = 100000,
+    ) -> "ChessTokenizer":
+        """
+        Build a tokenizer vocabulary from a Hugging Face dataset.
+        
+        Args:
+            dataset_name: Name of the dataset on Hugging Face Hub.
+            split: Dataset split to use.
+            column: Column containing the game strings.
+            min_frequency: Minimum frequency for a token to be included (default: 500).
+            max_samples: Maximum number of samples to process (default: 100k).
+        
+        Returns:
+            A ChessTokenizer with the built vocabulary.
+        """
+        from datasets import load_dataset
+        
+        dataset = load_dataset(dataset_name, split=split)
+        
+        if max_samples is not None:
+            dataset = dataset.select(range(min(max_samples, len(dataset))))
+        
+        def game_iterator():
+            for example in dataset:
+                yield example[column]
+        
+        return cls.build_vocab_from_iterator(game_iterator(), min_frequency=min_frequency)
+    
+    @property
+    def vocab_size(self) -> int:
+        """Return the size of the vocabulary."""
+        return len(self._vocab)
+    
+    def get_vocab(self) -> Dict[str, int]:
+        """Return the vocabulary as a dictionary."""
+        return dict(self._vocab)
+    
+    # def _tokenize(self, text: str) -> List[str]:
+    #     """
+    #     Tokenize a string of moves into a list of tokens.
+        
+    #     Args:
+    #         text: A string of space-separated moves.
+        
+    #     Returns:
+    #         List of move tokens.
+    #     """
+    #     return text.strip().split()
+    def _tokenize(self, text: str) -> List[str]:
+        parts = text.strip().split()
+        out: List[str] = []
+        special = {self.PAD_TOKEN, self.BOS_TOKEN, self.EOS_TOKEN, self.UNK_TOKEN}
+
+        for p in parts:
+            if p in special:
+                out.append(p)
+            else:
+                out.extend(self._decompose_move(p))
+        return out
+
+
+    @classmethod
+    def build_structured_vocab(cls) -> "ChessTokenizer":
+        special = [cls.PAD_TOKEN, cls.BOS_TOKEN, cls.EOS_TOKEN, cls.UNK_TOKEN]
+
+        # 12 color+piece tokens
+        cp = [f"{c}{p}" for c in ("W", "B") for p in ("P", "N", "B", "R", "Q", "K")]
+
+        files = "abcdefgh"
+        ranks = "12345678"
+
+        from_tokens = [f"{f}{r}_f" for f in files for r in ranks]  # 64
+        to_tokens = [f"{f}{r}_t" for f in files for r in ranks]    # 64
+
+        suffix = ["cap", "check", "mate", "castle_k", "castle_q"]
+        promo = [f"promo_{p}" for p in ("q", "r", "b", "n")]
+
+        tokens = special + cp + from_tokens + to_tokens + suffix + promo
+        vocab = {t: i for i, t in enumerate(tokens)}
+        return cls(vocab=vocab)
+
+    def _convert_token_to_id(self, token: str) -> int:
+        """Convert a token to its ID."""
+        return self._vocab.get(token, self._vocab.get(self.UNK_TOKEN, 0))
+    
+    def _convert_id_to_token(self, index: int) -> str:
+        """Convert an ID to its token."""
+        return self._ids_to_tokens.get(index, self.UNK_TOKEN)
+    
+    def convert_tokens_to_string(self, tokens: List[str]) -> str:
+        """Convert a list of tokens back to a string."""
+        # Filter out special tokens for cleaner output
+        special = {self.PAD_TOKEN, self.BOS_TOKEN, self.EOS_TOKEN, self.UNK_TOKEN}
+        return " ".join(t for t in tokens if t not in special)
+    
+    def save_vocabulary(
+        self,
+        save_directory: str,
+        filename_prefix: Optional[str] = None,
+    ) -> tuple:
+        """
+        Save the vocabulary to a JSON file.
+        
+        Args:
+            save_directory: Directory to save the vocabulary.
+            filename_prefix: Optional prefix for the filename.
+        
+        Returns:
+            Tuple containing the path to the saved vocabulary file.
+        """
+        if not os.path.isdir(save_directory):
+            os.makedirs(save_directory, exist_ok=True)
+        
+        vocab_file = os.path.join(
+            save_directory,
+            (filename_prefix + "-" if filename_prefix else "") + "vocab.json",
+        )
+        
+        with open(vocab_file, "w", encoding="utf-8") as f:
+            json.dump(self._vocab, f, ensure_ascii=False, indent=2)
+        
+        return (vocab_file,)
+
+
+def count_vocab_from_dataset(
+    dataset_name: str = "dlouapre/lichess_2025-01_1M",
+    split: str = "train",
+    column: str = "text",
+    max_samples: Optional[int] = 10000,
+) -> Dict[str, int]:
+    """
+    Count token frequencies in a dataset (useful for vocabulary analysis).
+    
+    Args:
+        dataset_name: Name of the dataset on Hugging Face Hub.
+        split: Dataset split to use.
+        column: Column containing the game strings.
+        max_samples: Maximum number of samples to process.
+    
+    Returns:
+        Dictionary mapping tokens to their frequencies.
+    """
+    from collections import Counter
+    from datasets import load_dataset
+    
+    dataset = load_dataset(dataset_name, split=split)
+    
+    if max_samples is not None:
+        dataset = dataset.select(range(min(max_samples, len(dataset))))
+    
+    token_counts = Counter()
+    
+    for example in dataset:
+        moves = example[column].strip().split()
+        token_counts.update(moves)
+    
+    return dict(token_counts)

tokenizer_config.json

@@ -0,0 +1,50 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "[BOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "[EOS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenizer.ChessTokenizer",
+      null
+    ]
+  },
+  "bos_token": "[BOS]",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "[EOS]",
+  "extra_special_tokens": {},
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "tokenizer_class": "ChessTokenizer",
+  "unk_token": "[UNK]"
+}

vocab.json

@@ -0,0 +1,155 @@
+{
+  "[PAD]": 0,
+  "[BOS]": 1,
+  "[EOS]": 2,
+  "[UNK]": 3,
+  "WP": 4,
+  "WN": 5,
+  "WB": 6,
+  "WR": 7,
+  "WQ": 8,
+  "WK": 9,
+  "BP": 10,
+  "BN": 11,
+  "BB": 12,
+  "BR": 13,
+  "BQ": 14,
+  "BK": 15,
+  "a1_f": 16,
+  "a2_f": 17,
+  "a3_f": 18,
+  "a4_f": 19,
+  "a5_f": 20,
+  "a6_f": 21,
+  "a7_f": 22,
+  "a8_f": 23,
+  "b1_f": 24,
+  "b2_f": 25,
+  "b3_f": 26,
+  "b4_f": 27,
+  "b5_f": 28,
+  "b6_f": 29,
+  "b7_f": 30,
+  "b8_f": 31,
+  "c1_f": 32,
+  "c2_f": 33,
+  "c3_f": 34,
+  "c4_f": 35,
+  "c5_f": 36,
+  "c6_f": 37,
+  "c7_f": 38,
+  "c8_f": 39,
+  "d1_f": 40,
+  "d2_f": 41,
+  "d3_f": 42,
+  "d4_f": 43,
+  "d5_f": 44,
+  "d6_f": 45,
+  "d7_f": 46,
+  "d8_f": 47,
+  "e1_f": 48,
+  "e2_f": 49,
+  "e3_f": 50,
+  "e4_f": 51,
+  "e5_f": 52,
+  "e6_f": 53,
+  "e7_f": 54,
+  "e8_f": 55,
+  "f1_f": 56,
+  "f2_f": 57,
+  "f3_f": 58,
+  "f4_f": 59,
+  "f5_f": 60,
+  "f6_f": 61,
+  "f7_f": 62,
+  "f8_f": 63,
+  "g1_f": 64,
+  "g2_f": 65,
+  "g3_f": 66,
+  "g4_f": 67,
+  "g5_f": 68,
+  "g6_f": 69,
+  "g7_f": 70,
+  "g8_f": 71,
+  "h1_f": 72,
+  "h2_f": 73,
+  "h3_f": 74,
+  "h4_f": 75,
+  "h5_f": 76,
+  "h6_f": 77,
+  "h7_f": 78,
+  "h8_f": 79,
+  "a1_t": 80,
+  "a2_t": 81,
+  "a3_t": 82,
+  "a4_t": 83,
+  "a5_t": 84,
+  "a6_t": 85,
+  "a7_t": 86,
+  "a8_t": 87,
+  "b1_t": 88,
+  "b2_t": 89,
+  "b3_t": 90,
+  "b4_t": 91,
+  "b5_t": 92,
+  "b6_t": 93,
+  "b7_t": 94,
+  "b8_t": 95,
+  "c1_t": 96,
+  "c2_t": 97,
+  "c3_t": 98,
+  "c4_t": 99,
+  "c5_t": 100,
+  "c6_t": 101,
+  "c7_t": 102,
+  "c8_t": 103,
+  "d1_t": 104,
+  "d2_t": 105,
+  "d3_t": 106,
+  "d4_t": 107,
+  "d5_t": 108,
+  "d6_t": 109,
+  "d7_t": 110,
+  "d8_t": 111,
+  "e1_t": 112,
+  "e2_t": 113,
+  "e3_t": 114,
+  "e4_t": 115,
+  "e5_t": 116,
+  "e6_t": 117,
+  "e7_t": 118,
+  "e8_t": 119,
+  "f1_t": 120,
+  "f2_t": 121,
+  "f3_t": 122,
+  "f4_t": 123,
+  "f5_t": 124,
+  "f6_t": 125,
+  "f7_t": 126,
+  "f8_t": 127,
+  "g1_t": 128,
+  "g2_t": 129,
+  "g3_t": 130,
+  "g4_t": 131,
+  "g5_t": 132,
+  "g6_t": 133,
+  "g7_t": 134,
+  "g8_t": 135,
+  "h1_t": 136,
+  "h2_t": 137,
+  "h3_t": 138,
+  "h4_t": 139,
+  "h5_t": 140,
+  "h6_t": 141,
+  "h7_t": 142,
+  "h8_t": 143,
+  "cap": 144,
+  "check": 145,
+  "mate": 146,
+  "castle_k": 147,
+  "castle_q": 148,
+  "promo_q": 149,
+  "promo_r": 150,
+  "promo_b": 151,
+  "promo_n": 152
+}