created datasets

.gitattributes

@@ -53,3 +53,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.jpg filter=lfs diff=lfs merge=lfs -text
 *.jpeg filter=lfs diff=lfs merge=lfs -text
 *.webp filter=lfs diff=lfs merge=lfs -text
+*.txt filter=lfs diff=lfs merge=lfs -text

chess_game_data.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f596fca8a3177bc5832e7a0f1e78fd14c3a924c53c3a15f4944a5d7ad869d1
+size 168853001

clean_data.ipynb

@@ -0,0 +1,304 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "import chess\n",
+    "import chess.pgn\n",
+    "import os\n",
+    "from tqdm import tqdm\n",
+    "import json"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      " 98%|█████████▊| 172350/175000 [38:39<00:35, 74.29it/s] \n"
+     ]
+    },
+    {
+     "ename": "UnicodeDecodeError",
+     "evalue": "'utf-8' codec can't decode byte 0xfc in position 3540: invalid start byte",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mUnicodeDecodeError\u001b[0m                        Traceback (most recent call last)",
+      "\u001b[1;32ms:\\Documents\\chess_engine\\clean_data.ipynb Cell 2\u001b[0m line \u001b[0;36m1\n\u001b[0;32m     <a href='vscode-notebook-cell:/s%3A/Documents/chess_engine/clean_data.ipynb#W1sZmlsZQ%3D%3D?line=12'>13</a>\u001b[0m     games\u001b[39m.\u001b[39mappend(game\u001b[39m.\u001b[39maccept(exporter)\u001b[39m.\u001b[39mreplace(\u001b[39m\"\u001b[39m\u001b[39m\\n\u001b[39;00m\u001b[39m\"\u001b[39m, \u001b[39m\"\u001b[39m\u001b[39m \u001b[39m\u001b[39m\"\u001b[39m) \u001b[39m+\u001b[39m \u001b[39m'\u001b[39m\u001b[39m\\n\u001b[39;00m\u001b[39m'\u001b[39m)\n\u001b[0;32m     <a href='vscode-notebook-cell:/s%3A/Documents/chess_engine/clean_data.ipynb#W1sZmlsZQ%3D%3D?line=13'>14</a>\u001b[0m     count \u001b[39m+\u001b[39m\u001b[39m=\u001b[39m \u001b[39m1\u001b[39m\n\u001b[1;32m---> <a href='vscode-notebook-cell:/s%3A/Documents/chess_engine/clean_data.ipynb#W1sZmlsZQ%3D%3D?line=14'>15</a>\u001b[0m     game \u001b[39m=\u001b[39m chess\u001b[39m.\u001b[39;49mpgn\u001b[39m.\u001b[39;49mread_game(game_file)\n\u001b[0;32m     <a href='vscode-notebook-cell:/s%3A/Documents/chess_engine/clean_data.ipynb#W1sZmlsZQ%3D%3D?line=15'>16</a>\u001b[0m     pbar\u001b[39m.\u001b[39mupdate(\u001b[39m1\u001b[39m)\n\u001b[0;32m     <a href='vscode-notebook-cell:/s%3A/Documents/chess_engine/clean_data.ipynb#W1sZmlsZQ%3D%3D?line=16'>17</a>\u001b[0m \u001b[39mwith\u001b[39;00m \u001b[39mopen\u001b[39m(\u001b[39m\"\u001b[39m\u001b[39mchess_game_data.txt\u001b[39m\u001b[39m\"\u001b[39m, \u001b[39m'\u001b[39m\u001b[39ma\u001b[39m\u001b[39m'\u001b[39m) \u001b[39mas\u001b[39;00m data_file:\n",
+      "File \u001b[1;32ms:\\packages\\envs\\chess\\Lib\\site-packages\\chess\\pgn.py:1673\u001b[0m, in \u001b[0;36mread_game\u001b[1;34m(handle, Visitor)\u001b[0m\n\u001b[0;32m   1671\u001b[0m \u001b[39mwhile\u001b[39;00m line \u001b[39mand\u001b[39;00m \u001b[39m\"\u001b[39m\u001b[39m}\u001b[39m\u001b[39m\"\u001b[39m \u001b[39mnot\u001b[39;00m \u001b[39min\u001b[39;00m line:\n\u001b[0;32m   1672\u001b[0m     comment_lines\u001b[39m.\u001b[39mappend(line)\n\u001b[1;32m-> 1673\u001b[0m     line \u001b[39m=\u001b[39m handle\u001b[39m.\u001b[39mreadline()\n\u001b[0;32m   1675\u001b[0m \u001b[39mif\u001b[39;00m line:\n\u001b[0;32m   1676\u001b[0m     close_index \u001b[39m=\u001b[39m line\u001b[39m.\u001b[39mfind(\u001b[39m\"\u001b[39m\u001b[39m}\u001b[39m\u001b[39m\"\u001b[39m)\n",
+      "File \u001b[1;32m<frozen codecs>:322\u001b[0m, in \u001b[0;36mdecode\u001b[1;34m(self, input, final)\u001b[0m\n",
+      "\u001b[1;31mUnicodeDecodeError\u001b[0m: 'utf-8' codec can't decode byte 0xfc in position 3540: invalid start byte"
+     ]
+    }
+   ],
+   "source": [
+    "count = 0\n",
+    "last = 0\n",
+    "with tqdm(total=175000) as pbar:\n",
+    "    for root, d_name, files in os.walk(\"S:\\Documents\\ccc-data\\ccc-data\"):\n",
+    "        for file in files:\n",
+    "            if os.path.splitext(file)[1] == '.pgn':\n",
+    "                pgn_path = os.path.join(root, file)\n",
+    "                with open(pgn_path) as game_file:\n",
+    "                    games = []\n",
+    "                    game = chess.pgn.read_game(game_file)\n",
+    "                    while game:\n",
+    "                        exporter = chess.pgn.StringExporter(headers=False, variations=False, comments=False)\n",
+    "                        games.append(game.accept(exporter).replace(\"\\n\", \" \") + '\\n')\n",
+    "                        count += 1\n",
+    "                        game = chess.pgn.read_game(game_file)\n",
+    "                        pbar.update(1)\n",
+    "                    with open(\"chess_game_data.txt\", 'a') as data_file:\n",
+    "                        data_file.writelines(games)\n",
+    "\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "  0%|          | 28/1000000 [00:00<2:01:47, 136.85it/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2021-12.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      " 49%|████▉     | 490442/1000000 [51:38<2:35:36, 54.58it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-01.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      " 97%|█████████▋| 970547/1000000 [1:39:46<09:18, 52.71it/s]   "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-02.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "1345189it [2:17:35, 63.03it/s]                             "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-03.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "1711576it [2:54:32, 51.88it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-04.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2050605it [3:28:36, 60.10it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-05.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2406133it [4:04:30, 54.89it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-06.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2741369it [4:38:51, 53.44it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-08.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "3087892it [5:15:18, 50.95it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-09.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "3434038it [5:53:10, 51.47it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-10.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "3761579it [6:28:19, 67.38it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-11.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "4064728it [7:00:58, 61.31it/s] "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "lichess_elite_2022-12.pgn\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "4343360it [7:29:28, 161.05it/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "count = 0\n",
+    "last = 0\n",
+    "with tqdm(total=1000000) as pbar:\n",
+    "    for root, d_name, files in os.walk(\"S:\\Documents\\Lichess Elite Database\\Latest\"):\n",
+    "        for file in files:\n",
+    "            print(file)\n",
+    "            if os.path.splitext(file)[1] == '.pgn':\n",
+    "                pgn_path = os.path.join(root, file)\n",
+    "                with open(pgn_path) as game_file:\n",
+    "                    games = []\n",
+    "                    try:\n",
+    "                        game = chess.pgn.read_game(game_file)\n",
+    "                        while game:\n",
+    "                            exporter = chess.pgn.StringExporter(headers=False, variations=False, comments=False)\n",
+    "                            games.append(game.accept(exporter).replace(\"\\n\", \" \") + '\\n')\n",
+    "                            count += 1\n",
+    "                            game = chess.pgn.read_game(game_file)\n",
+    "                            pbar.update(1)\n",
+    "                    except Exception as e:\n",
+    "                        print(e)\n",
+    "\n",
+    "                    with open(\"lichess_game_data.txt\", 'a') as data_file:\n",
+    "                        data_file.writelines(games)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "chess",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  },
+  "orig_nbformat": 4
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

lichess_game_data.txt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fbb174f8789c9e7c4eae50f6a2bf19a3757aa13788182a3cd8c61acb039d2e97
+size 2249268189

test_dataset.ipynb

@@ -0,0 +1,673 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 122,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset\n",
+    "import tokenize\n",
+    "from tokenizers import SentencePieceBPETokenizer, pre_tokenizers, ByteLevelBPETokenizer, normalizers\n",
+    "from transformers import AutoTokenizer"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 132,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Downloading data files: 100%|██████████| 1/1 [00:00<?, ?it/s]\n",
+      "Extracting data files: 100%|██████████| 1/1 [00:00<00:00, 111.41it/s]\n",
+      "Generating train split: 4343360 examples [00:24, 178753.60 examples/s]\n"
+     ]
+    }
+   ],
+   "source": [
+    "datasets = load_dataset(\"text\", data_files={\"train\": \"lichess_game_data.txt\"})"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 133,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def get_training_corpus():\n",
+    "    dataset = datasets[\"train\"]\n",
+    "    for start_idx in range(0, len(dataset), 1000):\n",
+    "        samples = dataset[start_idx : start_idx + 1000]\n",
+    "        yield samples['text']"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 134,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tokenizer = SentencePieceBPETokenizer()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 135,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tokenizer.normalizer = normalizers.Sequence([normalizers.Replace(\".\", \"\"), normalizers.Replace(\"1-0\", \"\"), normalizers.Replace(\"0-1\", \"\"), normalizers.Replace(\"1/2-1/2\", \"\")])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 136,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tokenizer.pre_tokenizer = pre_tokenizers.Split(\" \", 'removed')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 137,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "tokenizer.train_from_iterator(get_training_corpus(), vocab_size=10000000, min_frequency=0, show_progress=True, special_tokens=[\"<s>\", \"<pad>\", \"</s>\", \"<unk>\", \"<cls>\", \"<sep>\", \"<mask>\"])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 138,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "test_str = '1. e4 e5 2. Nf3 Nc6 3. d4 exd4 4. Nxd4 Nf6 5. Nxc6 bxc6 6. Bd3 d5 7. exd5 cxd5 8. O-O Be7 9. c4 O-O 10. cxd5 Bb7 11. Nc3 Nxd5 12. Bd2 Nb4 13. Be4 Bxe4 14. Nxe4 c5 15. Bc3 Qxd1 16. Rfxd1 Rad8 17. a3 Nd5 18. Ba5 Nb6 19. Nc3 Bf6 20. Rac1 Bxc3 21. Bxc3 Rxd1+ 22. Rxd1 f6 23. Kf1 Kf7 24. Rc1 Rd8 25. Ke2 Ke6 26. Bd2 c4 27. Be3 Rb8 28. Rc2 Rb7 29. h3 Kd5 30. Rd2+ Ke6 31. Rd4 a5 32. g3 Nd5 33. Bc1 Nb6 34. h4 g6 35. Bd2 Nd5 36. Bc1 Nb6 37. Bd2 Nd5 38. Rxc4 Rxb2 39. Ke1 Nb6 40. Re4+ Kd5 41. Re7 Rb1+ 42. Ke2 Nc4 43. Bc3 f5 44. Rd7+ Ke6 45. Rxh7 Rb3 46. Bd4 Rxa3 47. Rc7 Ra2+ 48. Kf3 Kd5 49. Rd7+ Kc6 50. Rg7 Kd5 51. Rd7+ Kc6 52. Rg7 Kd5 53. Rd7+ 1/2-1/2'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 139,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[('1', (0, 1)),\n",
+       " ('e4', (2, 4)),\n",
+       " ('e5', (5, 7)),\n",
+       " ('2', (8, 9)),\n",
+       " ('Nf3', (10, 13)),\n",
+       " ('Nc6', (14, 17)),\n",
+       " ('3', (18, 19)),\n",
+       " ('d4', (20, 22)),\n",
+       " ('exd4', (23, 27)),\n",
+       " ('4', (28, 29)),\n",
+       " ('Nxd4', (30, 34)),\n",
+       " ('Nf6', (35, 38)),\n",
+       " ('5', (39, 40)),\n",
+       " ('Nxc6', (41, 45)),\n",
+       " ('bxc6', (46, 50)),\n",
+       " ('6', (51, 52)),\n",
+       " ('Bd3', (53, 56)),\n",
+       " ('d5', (57, 59)),\n",
+       " ('7', (60, 61)),\n",
+       " ('exd5', (62, 66)),\n",
+       " ('cxd5', (67, 71)),\n",
+       " ('8', (72, 73)),\n",
+       " ('O-O', (74, 77)),\n",
+       " ('Be7', (78, 81)),\n",
+       " ('9', (82, 83)),\n",
+       " ('c4', (84, 86)),\n",
+       " ('O-O', (87, 90)),\n",
+       " ('10', (91, 93)),\n",
+       " ('cxd5', (94, 98)),\n",
+       " ('Bb7', (99, 102)),\n",
+       " ('11', (103, 105)),\n",
+       " ('Nc3', (106, 109)),\n",
+       " ('Nxd5', (110, 114)),\n",
+       " ('12', (115, 117)),\n",
+       " ('Bd2', (118, 121)),\n",
+       " ('Nb4', (122, 125)),\n",
+       " ('13', (126, 128)),\n",
+       " ('Be4', (129, 132)),\n",
+       " ('Bxe4', (133, 137)),\n",
+       " ('14', (138, 140)),\n",
+       " ('Nxe4', (141, 145)),\n",
+       " ('c5', (146, 148)),\n",
+       " ('15', (149, 151)),\n",
+       " ('Bc3', (152, 155)),\n",
+       " ('Qxd1', (156, 160)),\n",
+       " ('16', (161, 163)),\n",
+       " ('Rfxd1', (164, 169)),\n",
+       " ('Rad8', (170, 174)),\n",
+       " ('17', (175, 177)),\n",
+       " ('a3', (178, 180)),\n",
+       " ('Nd5', (181, 184)),\n",
+       " ('18', (185, 187)),\n",
+       " ('Ba5', (188, 191)),\n",
+       " ('Nb6', (192, 195)),\n",
+       " ('19', (196, 198)),\n",
+       " ('Nc3', (199, 202)),\n",
+       " ('Bf6', (203, 206)),\n",
+       " ('20', (207, 209)),\n",
+       " ('Rac1', (210, 214)),\n",
+       " ('Bxc3', (215, 219)),\n",
+       " ('21', (220, 222)),\n",
+       " ('Bxc3', (223, 227)),\n",
+       " ('Rxd1+', (228, 233)),\n",
+       " ('22', (234, 236)),\n",
+       " ('Rxd1', (237, 241)),\n",
+       " ('f6', (242, 244)),\n",
+       " ('23', (245, 247)),\n",
+       " ('Kf1', (248, 251)),\n",
+       " ('Kf7', (252, 255)),\n",
+       " ('24', (256, 258)),\n",
+       " ('Rc1', (259, 262)),\n",
+       " ('Rd8', (263, 266)),\n",
+       " ('25', (267, 269)),\n",
+       " ('Ke2', (270, 273)),\n",
+       " ('Ke6', (274, 277)),\n",
+       " ('26', (278, 280)),\n",
+       " ('Bd2', (281, 284)),\n",
+       " ('c4', (285, 287)),\n",
+       " ('27', (288, 290)),\n",
+       " ('Be3', (291, 294)),\n",
+       " ('Rb8', (295, 298)),\n",
+       " ('28', (299, 301)),\n",
+       " ('Rc2', (302, 305)),\n",
+       " ('Rb7', (306, 309)),\n",
+       " ('29', (310, 312)),\n",
+       " ('h3', (313, 315)),\n",
+       " ('Kd5', (316, 319)),\n",
+       " ('30', (320, 322)),\n",
+       " ('Rd2+', (323, 327)),\n",
+       " ('Ke6', (328, 331)),\n",
+       " ('31', (332, 334)),\n",
+       " ('Rd4', (335, 338)),\n",
+       " ('a5', (339, 341)),\n",
+       " ('32', (342, 344)),\n",
+       " ('g3', (345, 347)),\n",
+       " ('Nd5', (348, 351)),\n",
+       " ('33', (352, 354)),\n",
+       " ('Bc1', (355, 358)),\n",
+       " ('Nb6', (359, 362)),\n",
+       " ('34', (363, 365)),\n",
+       " ('h4', (366, 368)),\n",
+       " ('g6', (369, 371)),\n",
+       " ('35', (372, 374)),\n",
+       " ('Bd2', (375, 378)),\n",
+       " ('Nd5', (379, 382)),\n",
+       " ('36', (383, 385)),\n",
+       " ('Bc1', (386, 389)),\n",
+       " ('Nb6', (390, 393)),\n",
+       " ('37', (394, 396)),\n",
+       " ('Bd2', (397, 400)),\n",
+       " ('Nd5', (401, 404)),\n",
+       " ('38', (405, 407)),\n",
+       " ('Rxc4', (408, 412)),\n",
+       " ('Rxb2', (413, 417)),\n",
+       " ('39', (418, 420)),\n",
+       " ('Ke1', (421, 424)),\n",
+       " ('Nb6', (425, 428)),\n",
+       " ('40', (429, 431)),\n",
+       " ('Re4+', (432, 436)),\n",
+       " ('Kd5', (437, 440)),\n",
+       " ('41', (441, 443)),\n",
+       " ('Re7', (444, 447)),\n",
+       " ('Rb1+', (448, 452)),\n",
+       " ('42', (453, 455)),\n",
+       " ('Ke2', (456, 459)),\n",
+       " ('Nc4', (460, 463)),\n",
+       " ('43', (464, 466)),\n",
+       " ('Bc3', (467, 470)),\n",
+       " ('f5', (471, 473)),\n",
+       " ('44', (474, 476)),\n",
+       " ('Rd7+', (477, 481)),\n",
+       " ('Ke6', (482, 485)),\n",
+       " ('45', (486, 488)),\n",
+       " ('Rxh7', (489, 493)),\n",
+       " ('Rb3', (494, 497)),\n",
+       " ('46', (498, 500)),\n",
+       " ('Bd4', (501, 504)),\n",
+       " ('Rxa3', (505, 509)),\n",
+       " ('47', (510, 512)),\n",
+       " ('Rc7', (513, 516)),\n",
+       " ('Ra2+', (517, 521)),\n",
+       " ('48', (522, 524)),\n",
+       " ('Kf3', (525, 528)),\n",
+       " ('Kd5', (529, 532)),\n",
+       " ('49', (533, 535)),\n",
+       " ('Rd7+', (536, 540)),\n",
+       " ('Kc6', (541, 544)),\n",
+       " ('50', (545, 547)),\n",
+       " ('Rg7', (548, 551)),\n",
+       " ('Kd5', (552, 555)),\n",
+       " ('51', (556, 558)),\n",
+       " ('Rd7+', (559, 563)),\n",
+       " ('Kc6', (564, 567)),\n",
+       " ('52', (568, 570)),\n",
+       " ('Rg7', (571, 574)),\n",
+       " ('Kd5', (575, 578)),\n",
+       " ('53', (579, 581)),\n",
+       " ('Rd7+', (582, 586))]"
+      ]
+     },
+     "execution_count": 139,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tokenizer.pre_tokenizer.pre_tokenize_str(tokenizer.normalizer.normalize_str(test_str))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 140,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "['1',\n",
+       " 'e4',\n",
+       " 'e5',\n",
+       " '2',\n",
+       " 'Nf3',\n",
+       " 'Nc6',\n",
+       " '3',\n",
+       " 'd4',\n",
+       " 'exd4',\n",
+       " '4',\n",
+       " 'Nxd4',\n",
+       " 'Nf6',\n",
+       " '5',\n",
+       " 'Nxc6',\n",
+       " 'bxc6',\n",
+       " '6',\n",
+       " 'Bd3',\n",
+       " 'd5',\n",
+       " '7',\n",
+       " 'exd5',\n",
+       " 'cxd5',\n",
+       " '8',\n",
+       " 'O-O',\n",
+       " 'Be7',\n",
+       " '9',\n",
+       " 'c4',\n",
+       " 'O-O',\n",
+       " '10',\n",
+       " 'cxd5',\n",
+       " 'Bb7',\n",
+       " '11',\n",
+       " 'Nc3',\n",
+       " 'Nxd5',\n",
+       " '12',\n",
+       " 'Bd2',\n",
+       " 'Nb4',\n",
+       " '13',\n",
+       " 'Be4',\n",
+       " 'Bxe4',\n",
+       " '14',\n",
+       " 'Nxe4',\n",
+       " 'c5',\n",
+       " '15',\n",
+       " 'Bc3',\n",
+       " 'Qxd1',\n",
+       " '16',\n",
+       " 'Rfxd1',\n",
+       " 'Rad8',\n",
+       " '17',\n",
+       " 'a3',\n",
+       " 'Nd5',\n",
+       " '18',\n",
+       " 'Ba5',\n",
+       " 'Nb6',\n",
+       " '19',\n",
+       " 'Nc3',\n",
+       " 'Bf6',\n",
+       " '20',\n",
+       " 'Rac1',\n",
+       " 'Bxc3',\n",
+       " '21',\n",
+       " 'Bxc3',\n",
+       " 'Rxd1+',\n",
+       " '22',\n",
+       " 'Rxd1',\n",
+       " 'f6',\n",
+       " '23',\n",
+       " 'Kf1',\n",
+       " 'Kf7',\n",
+       " '24',\n",
+       " 'Rc1',\n",
+       " 'Rd8',\n",
+       " '25',\n",
+       " 'Ke2',\n",
+       " 'Ke6',\n",
+       " '26',\n",
+       " 'Bd2',\n",
+       " 'c4',\n",
+       " '27',\n",
+       " 'Be3',\n",
+       " 'Rb8',\n",
+       " '28',\n",
+       " 'Rc2',\n",
+       " 'Rb7',\n",
+       " '29',\n",
+       " 'h3',\n",
+       " 'Kd5',\n",
+       " '30',\n",
+       " 'Rd2+',\n",
+       " 'Ke6',\n",
+       " '31',\n",
+       " 'Rd4',\n",
+       " 'a5',\n",
+       " '32',\n",
+       " 'g3',\n",
+       " 'Nd5',\n",
+       " '33',\n",
+       " 'Bc1',\n",
+       " 'Nb6',\n",
+       " '34',\n",
+       " 'h4',\n",
+       " 'g6',\n",
+       " '35',\n",
+       " 'Bd2',\n",
+       " 'Nd5',\n",
+       " '36',\n",
+       " 'Bc1',\n",
+       " 'Nb6',\n",
+       " '37',\n",
+       " 'Bd2',\n",
+       " 'Nd5',\n",
+       " '38',\n",
+       " 'Rxc4',\n",
+       " 'Rxb2',\n",
+       " '39',\n",
+       " 'Ke1',\n",
+       " 'Nb6',\n",
+       " '40',\n",
+       " 'Re4+',\n",
+       " 'Kd5',\n",
+       " '41',\n",
+       " 'Re7',\n",
+       " 'Rb1+',\n",
+       " '42',\n",
+       " 'Ke2',\n",
+       " 'Nc4',\n",
+       " '43',\n",
+       " 'Bc3',\n",
+       " 'f5',\n",
+       " '44',\n",
+       " 'Rd7+',\n",
+       " 'Ke6',\n",
+       " '45',\n",
+       " 'Rxh7',\n",
+       " 'Rb3',\n",
+       " '46',\n",
+       " 'Bd4',\n",
+       " 'Rxa3',\n",
+       " '47',\n",
+       " 'Rc7',\n",
+       " 'Ra2+',\n",
+       " '48',\n",
+       " 'Kf3',\n",
+       " 'Kd5',\n",
+       " '49',\n",
+       " 'Rd7+',\n",
+       " 'Kc6',\n",
+       " '50',\n",
+       " 'Rg7',\n",
+       " 'Kd5',\n",
+       " '51',\n",
+       " 'Rd7+',\n",
+       " 'Kc6',\n",
+       " '52',\n",
+       " 'Rg7',\n",
+       " 'Kd5',\n",
+       " '53',\n",
+       " 'Rd7+']"
+      ]
+     },
+     "execution_count": 140,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tokenizer.encode(test_str).tokens"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 142,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[11,\n",
+       " 48,\n",
+       " 53,\n",
+       " 12,\n",
+       " 79,\n",
+       " 116,\n",
+       " 13,\n",
+       " 45,\n",
+       " 307,\n",
+       " 14,\n",
+       " 193,\n",
+       " 74,\n",
+       " 15,\n",
+       " 434,\n",
+       " 391,\n",
+       " 16,\n",
+       " 154,\n",
+       " 46,\n",
+       " 17,\n",
+       " 190,\n",
+       " 199,\n",
+       " 18,\n",
+       " 51,\n",
+       " 150,\n",
+       " 19,\n",
+       " 65,\n",
+       " 51,\n",
+       " 71,\n",
+       " 199,\n",
+       " 229,\n",
+       " 76,\n",
+       " 102,\n",
+       " 244,\n",
+       " 78,\n",
+       " 253,\n",
+       " 405,\n",
+       " 80,\n",
+       " 342,\n",
+       " 345,\n",
+       " 82,\n",
+       " 228,\n",
+       " 63,\n",
+       " 83,\n",
+       " 400,\n",
+       " 832,\n",
+       " 84,\n",
+       " 1660,\n",
+       " 432,\n",
+       " 86,\n",
+       " 117,\n",
+       " 176,\n",
+       " 88,\n",
+       " 692,\n",
+       " 319,\n",
+       " 90,\n",
+       " 102,\n",
+       " 322,\n",
+       " 91,\n",
+       " 459,\n",
+       " 386,\n",
+       " 92,\n",
+       " 386,\n",
+       " 791,\n",
+       " 95,\n",
+       " 519,\n",
+       " 177,\n",
+       " 96,\n",
+       " 273,\n",
+       " 201,\n",
+       " 98,\n",
+       " 222,\n",
+       " 214,\n",
+       " 103,\n",
+       " 239,\n",
+       " 246,\n",
+       " 105,\n",
+       " 253,\n",
+       " 65,\n",
+       " 110,\n",
+       " 155,\n",
+       " 241,\n",
+       " 113,\n",
+       " 346,\n",
+       " 421,\n",
+       " 114,\n",
+       " 107,\n",
+       " 324,\n",
+       " 115,\n",
+       " 808,\n",
+       " 246,\n",
+       " 118,\n",
+       " 414,\n",
+       " 77,\n",
+       " 121,\n",
+       " 112,\n",
+       " 176,\n",
+       " 122,\n",
+       " 572,\n",
+       " 319,\n",
+       " 125,\n",
+       " 73,\n",
+       " 100,\n",
+       " 126,\n",
+       " 253,\n",
+       " 176,\n",
+       " 128,\n",
+       " 572,\n",
+       " 319,\n",
+       " 131,\n",
+       " 253,\n",
+       " 176,\n",
+       " 136,\n",
+       " 553,\n",
+       " 578,\n",
+       " 139,\n",
+       " 487,\n",
+       " 319,\n",
+       " 143,\n",
+       " 1064,\n",
+       " 324,\n",
+       " 145,\n",
+       " 373,\n",
+       " 854,\n",
+       " 149,\n",
+       " 239,\n",
+       " 254,\n",
+       " 153,\n",
+       " 400,\n",
+       " 140,\n",
+       " 156,\n",
+       " 785,\n",
+       " 246,\n",
+       " 160,\n",
+       " 820,\n",
+       " 475,\n",
+       " 165,\n",
+       " 305,\n",
+       " 708,\n",
+       " 169,\n",
+       " 326,\n",
+       " 801,\n",
+       " 179,\n",
+       " 225,\n",
+       " 324,\n",
+       " 186,\n",
+       " 785,\n",
+       " 350,\n",
+       " 194,\n",
+       " 595,\n",
+       " 324,\n",
+       " 200,\n",
+       " 785,\n",
+       " 350,\n",
+       " 210,\n",
+       " 595,\n",
+       " 324,\n",
+       " 224,\n",
+       " 785]"
+      ]
+     },
+     "execution_count": 142,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tokenizer.encode(test_str)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 143,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "13677"
+      ]
+     },
+     "execution_count": 143,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "tokenizer.get_vocab_size()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "chess",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.5"
+  },
+  "orig_nbformat": 4
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}