first commit

.gitignore

@@ -0,0 +1,45 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Virtual Environment
+venv/
+ENV/
+env/
+virtual/
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+.DS_Store?
+._*
+.Spotlight-V100
+.Trashes
+ehthumbs.db
+Thumbs.db
+
+# Project specific
+tok_*.txt

README.md

@@ -1,13 +1,51 @@
----
-title: Bpe Tok
-emoji: 🌍
-colorFrom: indigo
-colorTo: blue
-sdk: gradio
-sdk_version: 5.12.0
-app_file: app.py
-pinned: false
-short_description: A short custom,  multilingual impl. of BPE tokenizer.
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# BPE Tokenization with Custom Regex
+
+This notebook demonstrates the process of training a Byte Pair Encoding (BPE) tokenizer using a custom regex pattern. The tokenizer is designed to handle multilingual text, specifically English, Hindi, and Kannada. The notebook includes steps for tokenization, vocabulary building, and encoding/decoding of text.
+
+## Overview
+
+The notebook is structured into several key sections:
+
+1. **Custom Regex for BPE Tokenization**: 
+   - A custom regex pattern is defined to tokenize text in English, Hindi, and Kannada. This pattern is used to split text into tokens that are then processed by the BPE algorithm.
+
+2. **Dataset Loading**:
+   - Datasets from Hugging Face are loaded for English, Hindi (Devanagari script), and Kannada. These datasets are used to create a corpus for training the tokenizer.
+
+3. **Corpus Preparation**:
+   - Texts from the datasets are concatenated into a single corpus, which is then saved to a file. This corpus serves as the input for training the BPE tokenizer.
+
+4. **Utility Functions**:
+   - Functions are defined to handle control characters, visualize tokens, and manage token rendering.
+
+5. **Training BPE**:
+   - The BPE algorithm is trained on the prepared corpus. The process involves iteratively merging the most frequent pairs of tokens until the desired vocabulary size is reached.
+
+6. **Vocabulary and Model Saving**:
+   - The trained vocabulary and model are saved to disk for later use. The vocabulary consists of 3257 tokens, which includes:
+     - 256 byte-level tokens
+     - 3000 merge operations
+     - 1 special `<|endoftext|>` token
+
+7. **Encoding and Decoding**:
+   - Functions are provided to encode text into token IDs and decode token IDs back into text. Special tokens are handled as part of this process.
+
+8. **Testing**:
+   - The tokenizer is tested on sample texts to verify its performance and compression ratio.
+
+## Key Details
+
+- **Vocabulary Size**: The final vocabulary size is set to 3257 tokens (256 byte tokens + 3000 merges + 1 `<|endoftext|>` token).
+- **Tokenizer Training Corpus Composition**: The training corpus is constructed by combining texts from multiple datasets with the following distribution:
+  - `HuggingFaceFW/fineweb-edu` (English): 60% of the corpus, aligning with the token distribution patterns observed in advanced language models like GPT-4, where English tokens constitute a significant majority
+  - `ai4bharat/sangraha` (Hindi - Devanagari script): 20% of the corpus
+  - `ai4bharat/sangraha` (Kannada - Kannada script): 20% of the corpus
+- **Compression Ratio**: The compression ratio achieved by the BPE tokenizer is approximately 4.07x, indicating the efficiency of the tokenization process in reducing the size of the text representation.
+
+## Usage
+
+To use the tokenizer, load the saved model and vocabulary files, and utilize the provided encoding and decoding functions to process text. The tokenizer is capable of handling multilingual text and special tokens, making it suitable for diverse applications.
+
+## Conclusion
+
+This notebook provides a comprehensive guide to training a BPE tokenizer with custom regex patterns for multilingual text. The process includes dataset preparation, tokenization, vocabulary building, and model saving, offering a robust solution for text processing tasks.

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+import tiktoken
+from tokenizer import CustomTokenizer
+
+# Initialize tokenizers
+custom_tokenizer = CustomTokenizer("bpe_tok.model")
+tiktoken_encoder = tiktoken.get_encoding("gpt2")
+
+def encode_text(text):
+    # Get encodings from both tokenizers
+    custom_tokens = custom_tokenizer.encode(text, allowed_special={"<|endoftext|>"})
+    tiktoken_tokens = tiktoken_encoder.encode(text)
+    
+    # Format output
+    custom_output = f"Token count: {len(custom_tokens)}\nTokens: {custom_tokens}"
+    tiktoken_output = f"Token count: {len(tiktoken_tokens)}\nTokens: {tiktoken_tokens}"
+    
+    return custom_output, tiktoken_output
+
+# Create Gradio interface
+iface = gr.Interface(
+    fn=encode_text,
+    inputs=gr.Textbox(lines=5, label="Enter text to tokenize"),
+    outputs=[
+        gr.Textbox(label="Custom Tokenizer Output", lines=4),
+        gr.Textbox(label="Tiktoken Output", lines=4)
+    ],
+    title="Tokenizer Comparison",
+    description="Compare custom BPE tokenizer with Tiktoken GPT-2 tokenizer",
+    examples=[
+        ["आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।"],
+        ["मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together."],
+        ["My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>"]
+    ]
+)
+
+if __name__ == "__main__":
+    iface.launch() 

bpe_tok.ipynb

@@ -0,0 +1,1750 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Custom regex for BPE tokenization"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# wanna handle en, hi, kn languages.\n",
+    "tests = [\n",
+    "    \"आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।\",\n",
+    "    \"मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together.\",\n",
+    "    \"स्वल्पा adjust करो, बैंगलोर का ट्रैफिक ऐसा ही है।\",\n",
+    "    \"ನೀವು ಚಹಾ ಕುಡಿತೀರಾ? मुझे एक cup चाहिए।\",\n",
+    "    \"आज का काम पूरा करो, ನಾಳೆ ಎಲ್ಲಿಂದ ಆರಂಭಿಸೋದು ನೋಡಿ।\",\n",
+    "    \"ಪಾರ್ಟಿ ಹೇಗೆ ಇತ್ತು? मुझे तो बहुत मजा आया!\",\n",
+    "    \"ನಮ್ಮ ಚೂರು ಸಹನಶೀಲತೆಯನ್ನು ತೋರಿಸಿ, ये थोड़ी देर का मसला है।\",\n",
+    "    \"ಸಮಯ ನಿಲ್ಲುತ್ತಿಲ್ಲ, जिंदगी में स्वल्पा मज़ा भी जरूरी है।\",\n",
+    "    \"My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>\"\n",
+    "]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।\n",
+      "['आज', ' त', 'ो', ' बह', 'ुत', ' थक', ' गय', 'ा', ' ह', 'ूँ,', ' ಸ', '್ವಲ', '್ಪ', ' व', 'िश', '्र', 'ಾಂ', 'ತ', 'ಿ', ' ಬ', 'ೇಕ', 'ು।']\n",
+      "\n",
+      "मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together.\n",
+      "['म', 'ौसम', ' क', 'ितन', 'ा', ' अच', '्छ', 'ा', ' ह', 'ै!', ' ನ', 'ೀವ', 'ೂ', ' ಹ', 'ೊರಗ', 'ೆ', ' ಬನ', '್ನ', 'ಿ,', ' let', \"'s\", ' enjoy', ' together', '.']\n",
+      "\n",
+      "स्वल्पा adjust करो, बैंगलोर का ट्रैफिक ऐसा ही है।\n",
+      "['स', '्वल', '्प', 'ा', ' adjust', ' कर', 'ो,', ' ब', 'ैं', 'गल', 'ोर', ' क', 'ा', ' ट', '्र', 'ैफ', 'िक', ' ऐस', 'ा', ' ह', 'ी', ' ह', 'ै।']\n",
+      "\n",
+      "ನೀವು ಚಹಾ ಕುಡಿತೀರಾ? मुझे एक cup चाहिए।\n",
+      "['ನ', 'ೀವ', 'ು', ' ಚಹ', 'ಾ', ' ಕ', 'ುಡ', 'ಿತ', 'ೀರ', 'ಾ?', ' म', 'ुझ', 'े', ' एक', ' cup', ' च', 'ाह', 'िए', '।']\n",
+      "\n",
+      "आज का काम पूरा करो, ನಾಳೆ ಎಲ್ಲಿಂದ ಆರಂಭಿಸೋದು ನೋಡಿ।\n",
+      "['आज', ' क', 'ा', ' क', 'ाम', ' प', 'ूर', 'ा', ' कर', 'ो,', ' ನ', 'ಾಳ', 'ೆ', ' ಎಲ', '್ಲ', 'ಿಂ', 'ದ', ' ಆರ', 'ಂಭ', 'ಿಸ', 'ೋದ', 'ು', ' ನ', 'ೋಡ', 'ಿ।']\n",
+      "\n",
+      "ಪಾರ್ಟಿ ಹೇಗೆ ಇತ್ತು? मुझे तो बहुत मजा आया!\n",
+      "['ಪ', 'ಾರ', '್ಟ', 'ಿ', ' ಹ', 'ೇಗ', 'ೆ', ' ಇತ', '್ತ', 'ು?', ' म', 'ुझ', 'े', ' त', 'ो', ' बह', 'ुत', ' मज', 'ा', ' आय', 'ा!']\n",
+      "\n",
+      "ನಮ್ಮ ಚೂರು ಸಹನಶೀಲತೆಯನ್ನು ತೋರಿಸಿ, ये थोड़ी देर का मसला है।\n",
+      "['ನಮ', '್ಮ', ' ಚ', 'ೂರ', 'ು', ' ಸಹನಶ', 'ೀಲತ', 'ೆಯನ', '್ನ', 'ು', ' ತ', 'ೋರ', 'ಿಸ', 'ಿ,', ' य', 'े', ' थ', 'ोड', '़ी', ' द', 'ेर', ' क', 'ा', ' मसल', 'ा', ' ह', 'ै।']\n",
+      "\n",
+      "ಸಮಯ ನಿಲ್ಲುತ್ತಿಲ್ಲ, जिंदगी में स्वल्पा मज़ा भी जरूरी है।\n",
+      "['ಸಮಯ', ' ನ', 'ಿಲ', '್ಲ', 'ುತ', '್ತ', 'ಿಲ', '್ಲ', ',', ' ज', 'िं', 'दग', 'ी', ' म', 'ें', ' स', '्वल', '्प', 'ा', ' मज', '़ा', ' भ', 'ी', ' जर', 'ूर', 'ी', ' ह', 'ै।']\n",
+      "\n",
+      "My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>\n",
+      "['My', ' name', ' is', ' Jeff', ' Bezos', ',', ' and', ' I', \"'m\", ' the', ' owner', ' of', ' Amazon', '.<|', 'endoftext', '|>']\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "import regex as re\n",
+    "\n",
+    "def test_regex(regex, text):\n",
+    "    print(text)\n",
+    "    print(re.findall(regex, text))\n",
+    "    print()\n",
+    "\n",
+    "# default gpt4 regex\n",
+    "pattern = r\"\"\"'(?i:[sdmt]|ll|ve|re)|[^\\r\\n\\p{L}\\p{N}]?+\\p{L}+|\\p{N}{1,3}| ?[^\\s\\p{L}\\p{N}]++[\\r\\n]*|\\s*[\\r\\n]|\\s+(?!\\S)|\\s+\"\"\"\n",
+    "regex = re.compile(pattern)\n",
+    "\n",
+    "for test in tests:\n",
+    "    test_regex(regex, test)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।\n",
+      "['आज', ' तो', ' बहुत', ' थक', ' गया', ' हूँ', ',', ' ಸ್ವಲ್ಪ', ' ಬೇಕು', '।']\n",
+      "\n",
+      "मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together.\n",
+      "['मौसम', ' कितना', ' अच्छा', ' है', '!', ' ನೀವೂ', ' ಹೊರಗೆ', ' ಬನ್ನಿ', ',', ' let', \"'\", 's', ' enjoy', ' together', '.']\n",
+      "\n",
+      "स्वल्पा adjust करो, बैंगलोर का ट्रैफिक ऐसा ही है।\n",
+      "['स्वल्पा', ' adjust', ' करो', ',', ' बैंगलोर', ' का', ' ट्रैफिक', ' ऐसा', ' ही', ' है', '।']\n",
+      "\n",
+      "ನೀವು ಚಹಾ ಕುಡಿತೀರಾ? मुझे एक cup चाहिए।\n",
+      "['ನೀವು', ' ಚಹಾ', ' ಕುಡಿತೀರಾ', '?', ' मुझे', ' एक', ' cup', ' चाहिए', '।']\n",
+      "\n",
+      "आज का काम पूरा करो, ನಾಳೆ ಎಲ್ಲಿಂದ ಆರಂಭಿಸೋದು ನೋಡಿ।\n",
+      "['आज', ' का', ' काम', ' पूरा', ' करो', ',', ' ನಾಳೆ', ' ಎಲ್ಲಿಂದ', ' ಆರಂಭಿಸೋದು', ' ನೋಡಿ', '।']\n",
+      "\n",
+      "ಪಾರ್ಟಿ ಹೇಗೆ ಇತ್ತು? मुझे तो बहुत मजा आया!\n",
+      "['ಪಾರ್ಟಿ', ' ಹೇಗೆ', ' ಇತ್ತು', '?', ' मुझे', ' तो', ' बहुत', ' मजा', ' आया', '!']\n",
+      "\n",
+      "ನಮ್ಮ ಚೂರು ಸಹನಶೀಲತೆಯನ್ನು ತೋರಿಸಿ, ये थोड़ी देर का मसला है।\n",
+      "['ನಮ್ಮ', ' ಚೂರು', ' ಸಹನಶೀಲತೆಯನ್ನು', ' ತೋರಿಸಿ', ',', ' ये', ' थोड़ी', ' देर', ' का', ' मसला', ' है', '।']\n",
+      "\n",
+      "ಸಮಯ ನಿಲ್ಲುತ್ತಿಲ್ಲ, जिंदगी में स्वल्पा मज़ा भी जरूरी है।\n",
+      "['ಸಮಯ', ' ನಿಲ್ಲುತ್ತಿಲ್ಲ', ',', ' जिंदगी', ' में', ' स्वल्पा', ' मज़ा', ' भी', ' जरूरी', ' है', '।']\n",
+      "\n",
+      "My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>\n",
+      "['My', ' name', ' is', ' Jeff', ' Bezos', ',', ' and', ' I', \"'m\", ' the', ' owner', ' of', ' Amazon', '.', '<|', 'endoftext', '|>']\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "# custom regex\n",
+    "pattern = r\"\"\"(?i) 's|'t|'re|'ve|'m|'ll|'d| ?\\b[\\p{L}\\u0900-\\u0963|\\u0966-\\u097F]+\\b| ?\\b[\\p{L}\\u0C80-\\u0C9E|\\u0CA0-\\u0CFF]+\\b| ?[\\p{N}]+| ?[.,!?;:'\\\"-]| ?[\\u0964-\\u0965]| ?[\\u0C9E-\\u0C9F]| ?[^\\s\\p{L}\\p{N}\\u0900-\\u097F\\u0C80-\\u0CFF]+| \\s+(?!\\S)| \\s+\"\"\"\n",
+    "regex = re.compile(pattern)\n",
+    "\n",
+    "for test in tests:\n",
+    "    test_regex(regex, test)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Dataset"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset, load_from_disk\n",
+    "import random\n",
+    "import os\n",
+    "\n",
+    "# Load a subset of a dataset and cache it locally\n",
+    "def download_dataset_and_load_subset(dataset_name, name=None, data_dir=None, split=\"train\", num_rows=1000, save_dir=\"/home/yaseen/hf_datasets\"):\n",
+    "    # Create a directory path that includes the dataset name and split\n",
+    "    dataset_path = f\"{dataset_name.replace('/', '_')}_{name.replace('/', '_') if name else ''}_{data_dir.replace('/', '_') if data_dir else ''}_{split}\"\n",
+    "    save_path = os.path.join(save_dir, dataset_path)\n",
+    "    \n",
+    "    try:\n",
+    "        dataset = load_from_disk(save_path)\n",
+    "        print(f\"Dataset {dataset_name} loaded from {save_path}..\")\n",
+    "    \n",
+    "    except FileNotFoundError:\n",
+    "        # Load the dataset and cache it in the specified directory\n",
+    "        dataset = load_dataset(\n",
+    "            dataset_name, \n",
+    "            name=name, \n",
+    "            data_dir=data_dir, \n",
+    "            split=split\n",
+    "        )\n",
+    "        \n",
+    "        # Ensure the base save directory exists\n",
+    "        os.makedirs(save_dir, exist_ok=True)\n",
+    "        \n",
+    "        # Save the dataset to disk\n",
+    "        dataset.save_to_disk(save_path)\n",
+    "        print(f\"Dataset {dataset_name} saved to {save_path}..\")\n",
+    "    \n",
+    "    # Select a random subset of the dataset\n",
+    "    indices = list(range(len(dataset)))\n",
+    "    random.shuffle(indices)\n",
+    "    selected_indices = indices[:num_rows]\n",
+    "    return dataset.select(selected_indices)\n",
+    "\n",
+    "# Load subsets of each dataset with caching\n",
+    "dataset_en = download_dataset_and_load_subset(\"HuggingFaceFW/fineweb-edu\", name=\"sample-10BT\", num_rows=600)\n",
+    "dataset_hin_deva = download_dataset_and_load_subset(\"ai4bharat/sangraha\", data_dir=\"synthetic/hin_Deva\", num_rows=200)\n",
+    "dataset_kan_knda = download_dataset_and_load_subset(\"ai4bharat/sangraha\", data_dir=\"synthetic/kan_Knda\", num_rows=200)\n",
+    "\n",
+    "# Print the first row of each subset to verify\n",
+    "print(dataset_en[0])\n",
+    "print(dataset_hin_deva[0])\n",
+    "print(dataset_kan_knda[0])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Concatenate all texts into a single list\n",
+    "all_texts = []\n",
+    "\n",
+    "# Collect texts from each dataset\n",
+    "en_texts = [doc[\"text\"].strip().replace(\"\\n\", \" \") for doc in dataset_en]\n",
+    "hin_deva_texts = [doc[\"text\"].strip().replace(\"\\n\", \" \") for doc in dataset_hin_deva]\n",
+    "kan_knda_texts = [doc[\"text\"].strip().replace(\"\\n\", \" \") for doc in dataset_kan_knda]\n",
+    "\n",
+    "# Add all texts to a single list\n",
+    "all_texts.extend(en_texts)\n",
+    "all_texts.extend(hin_deva_texts)\n",
+    "all_texts.extend(kan_knda_texts)\n",
+    "\n",
+    "# Shuffle the combined texts\n",
+    "random.shuffle(all_texts)\n",
+    "\n",
+    "print(f\"Total number of texts: {len(all_texts)}\")\n",
+    "print(\"\\nFirst few texts after random shuffling:\")\n",
+    "print(all_texts[:3])\n",
+    "\n",
+    "corpus = \"\\n\".join(all_texts)\n",
+    "\n",
+    "# Save the combined texts to a file\n",
+    "with open(\"tok_corpus_small.txt\", \"w\", encoding=\"utf-8\") as file:\n",
+    "    file.write(corpus)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Utility functions"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import unicodedata\n",
+    "\n",
+    "def replace_control_characters(s: str) -> str:\n",
+    "    # we don't want to print control characters\n",
+    "    # which distort the output (e.g. \\n or much worse)\n",
+    "    # https://stackoverflow.com/questions/4324790/removing-control-characters-from-a-string-in-python/19016117#19016117\n",
+    "    # http://www.unicode.org/reports/tr44/#GC_Values_Table\n",
+    "    chars = []\n",
+    "    for ch in s:\n",
+    "        if unicodedata.category(ch)[0] != \"C\":\n",
+    "            chars.append(ch) # this character is ok\n",
+    "        else:\n",
+    "            chars.append(f\"\\\\u{ord(ch):04x}\") # escape\n",
+    "    return \"\".join(chars)\n",
+    "\n",
+    "def render_token(t: bytes) -> str:\n",
+    "    # pretty print a token, escaping control characters\n",
+    "    s = t.decode('utf-8', errors='replace')\n",
+    "    s = replace_control_characters(s)\n",
+    "    return s\n",
+    "\n",
+    "# utility function to visualise tokens\n",
+    "def visualise_tokens(token_values: list[bytes], end=\"\\n\") -> None:\n",
+    "    background = [f\"\\u001b[48;5;{i}m\" for i in [167, 179, 185, 77, 80, 68, 134]]\n",
+    "    # If token boundaries do not occur at unicode character boundaries, it's unclear how best to\n",
+    "    # visualise the token. Here, we'll just use the unicode replacement character to represent some\n",
+    "    # fraction of a character.\n",
+    "    unicode_token_values = [x.decode(\"utf-8\", errors=\"replace\") for x in token_values]\n",
+    "\n",
+    "    running_length = 0\n",
+    "    last_color = None\n",
+    "    for token in unicode_token_values:\n",
+    "        color = background[running_length % len(background)]\n",
+    "        if color == last_color:\n",
+    "            color = background[(running_length + 1) % len(background)]\n",
+    "            assert color != last_color\n",
+    "        last_color = color\n",
+    "        running_length += len(token)\n",
+    "        print(color + token, end=\"\")\n",
+    "    print(\"\\u001b[0m\", end=end)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Training BPE"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Corpus length in characters: 4,182,226\n",
+      "Corpus length in bytes: 6,164,138\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Read the corpus file\n",
+    "with open(\"tok_corpus_small.txt\", \"r\", encoding=\"utf-8\") as file:\n",
+    "    corpus = file.read()\n",
+    "\n",
+    "# Get character and byte lengths\n",
+    "char_length = len(corpus)\n",
+    "byte_length = len(corpus.encode('utf-8'))\n",
+    "\n",
+    "print(f\"Corpus length in characters: {char_length:,}\")\n",
+    "print(f\"Corpus length in bytes: {byte_length:,}\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "Training BPE: 100%|██████████| 3000/3000 [56:08<00:00,  1.12s/merge]  "
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "tokens length: 6142083\n",
+      "ids length: 1507403\n",
+      "compression ratio: 4.07X\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "from tqdm import tqdm\n",
+    "\n",
+    "def get_stats(ids, freq):\n",
+    "    for pair in zip(ids[:-1], ids[1:]):\n",
+    "        freq[pair] = freq.get(pair, 0) + 1\n",
+    "\n",
+    "def merge(ids, pair, idx):\n",
+    "  newids = []\n",
+    "  i = 0\n",
+    "  while i < len(ids):\n",
+    "    if i < len(ids) - 1 and ids[i] == pair[0] and ids[i+1] == pair[1]:\n",
+    "      newids.append(idx)\n",
+    "      i += 2\n",
+    "    else:\n",
+    "      newids.append(ids[i])\n",
+    "      i += 1\n",
+    "  return newids\n",
+    "\n",
+    "# ---\n",
+    "# tokens = corpus.encode('utf-8')\n",
+    "vocab_size = 3256 # the desired final vocabulary size\n",
+    "num_merges = vocab_size - 256\n",
+    "# ids = list(tokens) # copy so we don't destroy the original list\n",
+    "\n",
+    "text_chunks = re.findall(regex, corpus)\n",
+    "tokens = [list(ch.encode(\"utf-8\")) for ch in text_chunks]\n",
+    "ids = [list(ch.encode(\"utf-8\")) for ch in text_chunks]\n",
+    "\n",
+    "merges = {} # (int, int) -> int\n",
+    "vocab = {idx: bytes([idx]) for idx in range(256)}\n",
+    "\n",
+    "for i in tqdm(range(num_merges), desc=\"Training BPE\", unit=\"merge\"):\n",
+    "    stats = {}\n",
+    "    for chunk_ids in ids:\n",
+    "        get_stats(chunk_ids, stats)\n",
+    "    pair = max(stats, key=stats.get)\n",
+    "    idx = 256 + i\n",
+    "    ids = [merge(chunk_ids, pair, idx) for chunk_ids in ids]\n",
+    "    merges[pair] = idx\n",
+    "    vocab[idx] = vocab[pair[0]] + vocab[pair[1]]\n",
+    "\n",
+    "tok_len = 0\n",
+    "for chunk_toks in tokens:\n",
+    "    tok_len += len(chunk_toks)\n",
+    "print(\"tokens length:\", tok_len)\n",
+    "\n",
+    "ids_len = 0\n",
+    "for chunk_ids in ids:\n",
+    "    ids_len += len(chunk_ids)\n",
+    "print(\"ids length:\", ids_len)\n",
+    "\n",
+    "print(f\"compression ratio: {tok_len / ids_len:.2f}X\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def _build_vocab():\n",
+    "    vocab = {idx: bytes([idx]) for idx in range(256)}\n",
+    "    for (p0, p1), idx in merges.items():\n",
+    "        vocab[idx] = vocab[p0] + vocab[p1]\n",
+    "    return vocab\n",
+    "\n",
+    "special_tokens = {'<|endoftext|>': len(vocab)}\n",
+    "\n",
+    "def save(file_prefix):\n",
+    "    \"\"\"\n",
+    "    Saves two files: file_prefix.vocab and file_prefix.model\n",
+    "    This is inspired (but not equivalent to!) sentencepiece's model saving:\n",
+    "    - model file is the critical one, intended for load()\n",
+    "    - vocab file is just a pretty printed version for human inspection only\n",
+    "    \"\"\"\n",
+    "    # write the model: to be used in load() later\n",
+    "    model_file = file_prefix + \".model\"\n",
+    "    with open(model_file, 'w') as f:\n",
+    "        # write the version, pattern and merges, that's all that's needed\n",
+    "        f.write(\"minbpe v1\\n\")\n",
+    "        f.write(f\"{pattern}\\n\")\n",
+    "        # write the special tokens, first the number of them, then each one\n",
+    "        f.write(f\"{len(special_tokens)}\\n\")\n",
+    "        for special, idx in special_tokens.items():\n",
+    "            f.write(f\"{special} {idx}\\n\")\n",
+    "        # the merges dict\n",
+    "        for idx1, idx2 in merges:\n",
+    "            f.write(f\"{idx1} {idx2}\\n\")\n",
+    "    # write the vocab: for the human to look at\n",
+    "    vocab_file = file_prefix + \".vocab\"\n",
+    "    inverted_merges = {idx: pair for pair, idx in merges.items()}\n",
+    "    with open(vocab_file, \"w\", encoding=\"utf-8\") as f:\n",
+    "        for idx, token in vocab.items():\n",
+    "            # note: many tokens may be partial utf-8 sequences\n",
+    "            # and cannot be decoded into valid strings. Here we're using\n",
+    "            # errors='replace' to replace them with the replacement char.\n",
+    "            # this also means that we couldn't possibly use .vocab in load()\n",
+    "            # because decoding in this way is a lossy operation!\n",
+    "            s = render_token(token)\n",
+    "            # find the children of this token, if any\n",
+    "            if idx in inverted_merges:\n",
+    "                # if this token has children, render it nicely as a merge\n",
+    "                idx0, idx1 = inverted_merges[idx]\n",
+    "                s0 = render_token(vocab[idx0])\n",
+    "                s1 = render_token(vocab[idx1])\n",
+    "                f.write(f\"[{s0}][{s1}] -> [{s}] {idx}\\n\")\n",
+    "            else:\n",
+    "                # otherwise this is leaf token, just print it\n",
+    "                # (this should just be the first 256 tokens, the bytes)\n",
+    "                f.write(f\"[{s}] {idx}\\n\")\n",
+    "\n",
+    "save(\"bpe_tok\")\n",
+    "\n",
+    "def load(model_file):\n",
+    "    \"\"\"Inverse of save() but only for the model file\"\"\"\n",
+    "    assert model_file.endswith(\".model\")\n",
+    "    # read the model file\n",
+    "    merges = {}\n",
+    "    special_tokens = {}\n",
+    "    idx = 256\n",
+    "    with open(model_file, 'r', encoding=\"utf-8\") as f:\n",
+    "        # read the version\n",
+    "        version = f.readline().strip()\n",
+    "        assert version == \"minbpe v1\"\n",
+    "        # read the pattern\n",
+    "        pattern = f.readline().strip()\n",
+    "        # read the special tokens\n",
+    "        num_special = int(f.readline().strip())\n",
+    "        for _ in range(num_special):\n",
+    "            special, special_idx = f.readline().strip().split()\n",
+    "            special_tokens[special] = int(special_idx)\n",
+    "        # read the merges\n",
+    "        for line in f:\n",
+    "            idx1, idx2 = map(int, line.split())\n",
+    "            merges[(idx1, idx2)] = idx\n",
+    "            idx += 1\n",
+    "    vocab = _build_vocab()\n",
+    "    return merges, special_tokens, vocab\n",
+    "\n",
+    "merges, special_tokens, vocab = load(\"bpe_tok.model\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "�\n"
+     ]
+    }
+   ],
+   "source": [
+    "def decode(ids) -> str:\n",
+    "  part_bytes = []\n",
+    "  for id in ids:\n",
+    "      if id in vocab:\n",
+    "          part_bytes.append(vocab[id]) # id can be > 256 after merging\n",
+    "      elif id in special_tokens:\n",
+    "          part_bytes.append(special_tokens[id])\n",
+    "      else:\n",
+    "          raise ValueError(f\"id={id} not in vocab or special_tokens\")\n",
+    "  text_bytes = b\"\".join(part_bytes)\n",
+    "  text = text_bytes.decode(encoding=\"utf-8\", errors=\"replace\")\n",
+    "  return text\n",
+    "\n",
+    "print(decode([128]))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{(224, 178): 256,\n",
+       " (224, 164): 257,\n",
+       " (224, 179): 258,\n",
+       " (32, 257): 259,\n",
+       " (224, 165): 260,\n",
+       " (32, 256): 261,\n",
+       " (32, 116): 262,\n",
+       " (258, 141): 263,\n",
+       " (263, 256): 264,\n",
+       " (32, 97): 265,\n",
+       " (104, 101): 266,\n",
+       " (105, 110): 267,\n",
+       " (256, 191): 268,\n",
+       " (257, 190): 269,\n",
+       " (114, 101): 270,\n",
+       " (268, 256): 271,\n",
+       " (262, 266): 272,\n",
+       " (32, 111): 273,\n",
+       " (258, 129): 274,\n",
+       " (260, 135): 275,\n",
+       " (101, 114): 276,\n",
+       " (257, 176): 277,\n",
+       " (32, 115): 278,\n",
+       " (256, 190): 279,\n",
+       " (260, 141): 280,\n",
+       " (279, 256): 281,\n",
+       " (97, 116): 282,\n",
+       " (111, 110): 283,\n",
+       " (32, 119): 284,\n",
+       " (110, 100): 285,\n",
+       " (32, 99): 286,\n",
+       " (257, 191): 287,\n",
+       " (101, 110): 288,\n",
+       " (101, 115): 289,\n",
+       " (105, 115): 290,\n",
+       " (280, 257): 291,\n",
+       " (287, 257): 292,\n",
+       " (259, 149): 293,\n",
+       " (258, 134): 294,\n",
+       " (105, 116): 295,\n",
+       " (256, 176): 296,\n",
+       " (111, 114): 297,\n",
+       " (269, 257): 298,\n",
+       " (32, 98): 299,\n",
+       " (32, 102): 300,\n",
+       " (32, 112): 301,\n",
+       " (97, 110): 302,\n",
+       " (101, 100): 303,\n",
+       " (273, 102): 304,\n",
+       " (267, 103): 305,\n",
+       " (97, 108): 306,\n",
+       " (111, 117): 307,\n",
+       " (97, 114): 308,\n",
+       " (257, 130): 309,\n",
+       " (32, 267): 310,\n",
+       " (32, 109): 311,\n",
+       " (265, 285): 312,\n",
+       " (256, 166): 313,\n",
+       " (262, 111): 314,\n",
+       " (260, 128): 315,\n",
+       " (32, 100): 316,\n",
+       " (256, 168): 317,\n",
+       " (105, 99): 318,\n",
+       " (256, 178): 319,\n",
+       " (260, 139): 320,\n",
+       " (256, 181): 321,\n",
+       " (32, 104): 322,\n",
+       " (108, 101): 323,\n",
+       " (97, 115): 324,\n",
+       " (257, 168): 325,\n",
+       " (105, 283): 326,\n",
+       " (256, 164): 327,\n",
+       " (114, 111): 328,\n",
+       " (262, 104): 329,\n",
+       " (256, 130): 330,\n",
+       " (259, 184): 331,\n",
+       " (257, 149): 332,\n",
+       " (288, 116): 333,\n",
+       " (32, 101): 334,\n",
+       " (256, 151): 335,\n",
+       " (32, 108): 336,\n",
+       " (226, 128): 337,\n",
+       " (105, 108): 338,\n",
+       " (32, 84): 339,\n",
+       " (32, 270): 340,\n",
+       " (261, 174): 341,\n",
+       " (111, 109): 342,\n",
+       " (32, 110): 343,\n",
+       " (99, 116): 344,\n",
+       " (259, 174): 345,\n",
+       " (256, 179): 346,\n",
+       " (118, 101): 347,\n",
+       " (115, 116): 348,\n",
+       " (261, 184): 349,\n",
+       " (257, 164): 350,\n",
+       " (264, 168): 351,\n",
+       " (264, 164): 352,\n",
+       " (299, 101): 353,\n",
+       " (257, 178): 354,\n",
+       " (108, 121): 355,\n",
+       " (32, 290): 356,\n",
+       " (32, 103): 357,\n",
+       " (256, 149): 358,\n",
+       " (259, 170): 359,\n",
+       " (264, 176): 360,\n",
+       " (293, 275): 361,\n",
+       " (32, 73): 362,\n",
+       " (280, 277): 363,\n",
+       " (32, 49): 364,\n",
+       " (115, 101): 365,\n",
+       " (111, 116): 366,\n",
+       " (275, 309): 367,\n",
+       " (264, 178): 368,\n",
+       " (257, 184): 369,\n",
+       " (105, 109): 370,\n",
+       " (111, 119): 371,\n",
+       " (105, 100): 372,\n",
+       " (117, 116): 373,\n",
+       " (259, 185): 374,\n",
+       " (111, 108): 375,\n",
+       " (32, 65): 376,\n",
+       " (264, 175): 377,\n",
+       " (99, 101): 378,\n",
+       " (260, 136): 379,\n",
+       " (261, 133): 380,\n",
+       " (256, 161): 381,\n",
+       " (329, 282): 382,\n",
+       " (317, 351): 383,\n",
+       " (258, 135): 384,\n",
+       " (32, 83): 385,\n",
+       " (327, 352): 386,\n",
+       " (101, 116): 387,\n",
+       " (261, 170): 388,\n",
+       " (260, 164): 389,\n",
+       " (32, 117): 390,\n",
+       " (294, 256): 391,\n",
+       " (257, 174): 392,\n",
+       " (273, 110): 393,\n",
+       " (300, 297): 394,\n",
+       " (32, 67): 395,\n",
+       " (99, 104): 396,\n",
+       " (114, 97): 397,\n",
+       " (105, 103): 398,\n",
+       " (278, 116): 399,\n",
+       " (260, 129): 400,\n",
+       " (261, 149): 401,\n",
+       " (282, 326): 402,\n",
+       " (117, 114): 403,\n",
+       " (345, 367): 404,\n",
+       " (105, 114): 405,\n",
+       " (271, 184): 406,\n",
+       " (32, 121): 407,\n",
+       " (383, 274): 408,\n",
+       " (32, 266): 409,\n",
+       " (118, 276): 410,\n",
+       " (339, 266): 411,\n",
+       " (97, 100): 412,\n",
+       " (32, 40): 413,\n",
+       " (277, 291): 414,\n",
+       " (97, 121): 415,\n",
+       " (259, 156): 416,\n",
+       " (281, 176): 417,\n",
+       " (319, 368): 418,\n",
+       " (101, 108): 419,\n",
+       " (281, 151): 420,\n",
+       " (271, 175): 421,\n",
+       " (265, 115): 422,\n",
+       " (258, 138): 423,\n",
+       " (261, 168): 424,\n",
+       " (271, 166): 425,\n",
+       " (97, 109): 426,\n",
+       " (295, 104): 427,\n",
+       " (258, 139): 428,\n",
+       " (256, 184): 429,\n",
+       " (32, 50): 430,\n",
+       " (175, 269): 431,\n",
+       " (257, 170): 432,\n",
+       " (265, 270): 433,\n",
+       " (117, 108): 434,\n",
+       " (260, 130): 435,\n",
+       " (265, 110): 436,\n",
+       " (284, 427): 437,\n",
+       " (32, 295): 438,\n",
+       " (259, 181): 439,\n",
+       " (284, 104): 440,\n",
+       " (257, 185): 441,\n",
+       " (32, 77): 442,\n",
+       " (259, 172): 443,\n",
+       " (407, 307): 444,\n",
+       " (259, 143): 445,\n",
+       " (335, 346): 446,\n",
+       " (261, 181): 447,\n",
+       " (105, 102): 448,\n",
+       " (116, 276): 449,\n",
+       " (259, 168): 450,\n",
+       " (301, 328): 451,\n",
+       " (261, 185): 452,\n",
+       " (286, 283): 453,\n",
+       " (32, 80): 454,\n",
+       " (309, 257): 455,\n",
+       " (265, 108): 456,\n",
+       " (269, 277): 457,\n",
+       " (259, 133): 458,\n",
+       " (273, 114): 459,\n",
+       " (418, 268): 460,\n",
+       " (256, 159): 461,\n",
+       " (261, 172): 462,\n",
+       " (296, 274): 463,\n",
+       " (258, 128): 464,\n",
+       " (266, 114): 465,\n",
+       " (374, 379): 466,\n",
+       " (320, 309): 467,\n",
+       " (111, 100): 468,\n",
+       " (276, 115): 469,\n",
+       " (32, 87): 470,\n",
+       " (32, 66): 471,\n",
+       " (386, 274): 472,\n",
+       " (32, 72): 473,\n",
+       " (391, 175): 474,\n",
+       " (259, 166): 475,\n",
+       " (259, 148): 476,\n",
+       " (101, 109): 477,\n",
+       " (259, 178): 478,\n",
+       " (256, 174): 479,\n",
+       " (337, 153): 480,\n",
+       " (284, 101): 481,\n",
+       " (97, 285): 482,\n",
+       " (112, 101): 483,\n",
+       " (341, 472): 484,\n",
+       " (296, 264): 485,\n",
+       " (258, 130): 486,\n",
+       " (476, 277): 487,\n",
+       " (271, 164): 488,\n",
+       " (330, 313): 489,\n",
+       " (97, 103): 490,\n",
+       " (259, 176): 491,\n",
+       " (284, 324): 492,\n",
+       " (259, 151): 493,\n",
+       " (334, 120): 494,\n",
+       " (108, 100): 495,\n",
+       " (271, 149): 496,\n",
+       " (289, 115): 497,\n",
+       " (116, 104): 498,\n",
+       " (261, 134): 499,\n",
+       " (111, 112): 500,\n",
+       " (107, 101): 501,\n",
+       " (338, 108): 502,\n",
+       " (257, 151): 503,\n",
+       " (32, 69): 504,\n",
+       " (114, 105): 505,\n",
+       " (291, 175): 506,\n",
+       " (32, 337): 507,\n",
+       " (278, 117): 508,\n",
+       " (286, 342): 509,\n",
+       " (292, 149): 510,\n",
+       " (275, 257): 511,\n",
+       " (111, 270): 512,\n",
+       " (290, 116): 513,\n",
+       " (32, 118): 514,\n",
+       " (256, 170): 515,\n",
+       " (322, 97): 516,\n",
+       " (289, 116): 517,\n",
+       " (265, 116): 518,\n",
+       " (117, 109): 519,\n",
+       " (298, 168): 520,\n",
+       " (48, 48): 521,\n",
+       " (299, 121): 522,\n",
+       " (32, 34): 523,\n",
+       " (282, 101): 524,\n",
+       " (97, 98): 525,\n",
+       " (264, 159): 526,\n",
+       " (271, 151): 527,\n",
+       " (117, 115): 528,\n",
+       " (316, 101): 529,\n",
+       " (97, 267): 530,\n",
+       " (259, 165): 531,\n",
+       " (117, 270): 532,\n",
+       " (105, 347): 533,\n",
+       " (364, 57): 534,\n",
+       " (117, 110): 535,\n",
+       " (32, 68): 536,\n",
+       " (105, 289): 537,\n",
+       " (398, 104): 538,\n",
+       " (256, 172): 539,\n",
+       " (97, 99): 540,\n",
+       " (328, 109): 541,\n",
+       " (302, 116): 542,\n",
+       " (278, 101): 543,\n",
+       " (32, 76): 544,\n",
+       " (32, 82): 545,\n",
+       " (257, 181): 546,\n",
+       " (110, 116): 547,\n",
+       " (359, 363): 548,\n",
+       " (274, 321): 549,\n",
+       " (259, 137): 550,\n",
+       " (293, 320): 551,\n",
+       " (286, 104): 552,\n",
+       " (295, 121): 553,\n",
+       " (286, 302): 554,\n",
+       " (261, 142): 555,\n",
+       " (111, 115): 556,\n",
+       " (343, 366): 557,\n",
+       " (261, 176): 558,\n",
+       " (388, 360): 559,\n",
+       " (109, 333): 560,\n",
+       " (113, 117): 561,\n",
+       " (32, 70): 562,\n",
+       " (325, 275): 563,\n",
+       " (278, 104): 564,\n",
+       " (32, 78): 565,\n",
+       " (292, 431): 566,\n",
+       " (105, 118): 567,\n",
+       " (297, 116): 568,\n",
+       " (331, 275): 569,\n",
+       " (300, 541): 570,\n",
+       " (343, 101): 571,\n",
+       " (281, 168): 572,\n",
+       " (271, 168): 573,\n",
+       " (256, 175): 574,\n",
+       " (293, 315): 575,\n",
+       " (111, 99): 576,\n",
+       " (292, 164): 577,\n",
+       " (112, 112): 578,\n",
+       " (271, 176): 579,\n",
+       " (400, 257): 580,\n",
+       " (258, 136): 581,\n",
+       " (516, 347): 582,\n",
+       " (445, 332): 583,\n",
+       " (117, 100): 584,\n",
+       " (32, 114): 585,\n",
+       " (257, 175): 586,\n",
+       " (32, 71): 587,\n",
+       " (358, 264): 588,\n",
+       " (257, 172): 589,\n",
+       " (101, 344): 590,\n",
+       " (256, 163): 591,\n",
+       " (307, 495): 592,\n",
+       " (32, 45): 593,\n",
+       " (308, 116): 594,\n",
+       " (270, 115): 595,\n",
+       " (256, 185): 596,\n",
+       " (271, 130): 597,\n",
+       " (419, 108): 598,\n",
+       " (32, 323): 599,\n",
+       " (538, 116): 600,\n",
+       " (317, 263): 601,\n",
+       " (32, 107): 602,\n",
+       " (261, 135): 603,\n",
+       " (257, 188): 604,\n",
+       " (292, 175): 605,\n",
+       " (329, 290): 606,\n",
+       " (318, 306): 607,\n",
+       " (105, 306): 608,\n",
+       " (337, 157): 609,\n",
+       " (264, 166): 610,\n",
+       " (257, 159): 611,\n",
+       " (507, 156): 612,\n",
+       " (493, 257): 613,\n",
+       " (103, 101): 614,\n",
+       " (281, 166): 615,\n",
+       " (103, 104): 616,\n",
+       " (380, 321): 617,\n",
+       " (261, 164): 618,\n",
+       " (331, 291): 619,\n",
+       " (293, 269): 620,\n",
+       " (101, 270): 621,\n",
+       " (284, 297): 622,\n",
+       " (342, 101): 623,\n",
+       " (318, 104): 624,\n",
+       " (265, 98): 625,\n",
+       " (257, 161): 626,\n",
+       " (114, 307): 627,\n",
+       " (281, 175): 628,\n",
+       " (32, 79): 629,\n",
+       " (267, 101): 630,\n",
+       " (293, 277): 631,\n",
+       " (256, 182): 632,\n",
+       " (257, 166): 633,\n",
+       " (111, 103): 634,\n",
+       " (97, 112): 635,\n",
+       " (306, 108): 636,\n",
+       " (260, 137): 637,\n",
+       " (308, 100): 638,\n",
+       " (298, 178): 639,\n",
+       " (259, 134): 640,\n",
+       " (310, 116): 641,\n",
+       " (292, 143): 642,\n",
+       " (291, 164): 643,\n",
+       " (307, 116): 644,\n",
+       " (307, 114): 645,\n",
+       " (390, 115): 646,\n",
+       " (301, 108): 647,\n",
+       " (117, 348): 648,\n",
+       " (362, 110): 649,\n",
+       " (272, 405): 650,\n",
+       " (261, 151): 651,\n",
+       " (97, 107): 652,\n",
+       " (292, 178): 653,\n",
+       " (105, 112): 654,\n",
+       " (261, 156): 655,\n",
+       " (420, 268): 656,\n",
+       " (330, 327): 657,\n",
+       " (272, 121): 658,\n",
+       " (111, 348): 659,\n",
+       " (274, 386): 660,\n",
+       " (284, 266): 661,\n",
+       " (116, 465): 662,\n",
+       " (281, 130): 663,\n",
+       " (282, 303): 664,\n",
+       " (257, 156): 665,\n",
+       " (259, 164): 666,\n",
+       " (259, 135): 667,\n",
+       " (32, 370): 668,\n",
+       " (430, 48): 669,\n",
+       " (324, 116): 670,\n",
+       " (264, 151): 671,\n",
+       " (259, 182): 672,\n",
+       " (292, 184): 673,\n",
+       " (32, 85): 674,\n",
+       " (316, 111): 675,\n",
+       " (32, 74): 676,\n",
+       " (488, 274): 677,\n",
+       " (32, 288): 678,\n",
+       " (281, 178): 679,\n",
+       " (339, 104): 680,\n",
+       " (264, 181): 681,\n",
+       " (256, 156): 682,\n",
+       " (321, 408): 683,\n",
+       " (97, 270): 684,\n",
+       " (307, 115): 685,\n",
+       " (531, 269): 686,\n",
+       " (444, 114): 687,\n",
+       " (456, 108): 688,\n",
+       " (112, 116): 689,\n",
+       " (97, 344): 690,\n",
+       " (326, 115): 691,\n",
+       " (291, 181): 692,\n",
+       " (311, 101): 693,\n",
+       " (112, 108): 694,\n",
+       " (102, 102): 695,\n",
+       " (440, 624): 696,\n",
+       " (264, 184): 697,\n",
+       " (274, 313): 698,\n",
+       " (283, 103): 699,\n",
+       " (112, 276): 700,\n",
+       " (302, 115): 701,\n",
+       " (101, 308): 702,\n",
+       " (298, 174): 703,\n",
+       " (322, 290): 704,\n",
+       " (284, 502): 705,\n",
+       " (97, 501): 706,\n",
+       " (299, 373): 707,\n",
+       " (359, 277): 708,\n",
+       " (261, 146): 709,\n",
+       " (261, 166): 710,\n",
+       " (271, 178): 711,\n",
+       " (357, 111): 712,\n",
+       " (322, 324): 713,\n",
+       " (490, 101): 714,\n",
+       " (372, 101): 715,\n",
+       " (311, 512): 716,\n",
+       " (99, 108): 717,\n",
+       " (298, 164): 718,\n",
+       " (527, 294): 719,\n",
+       " (330, 381): 720,\n",
+       " (350, 269): 721,\n",
+       " (478, 642): 722,\n",
+       " (100, 117): 723,\n",
+       " (481, 270): 724,\n",
+       " (257, 154): 725,\n",
+       " (267, 100): 726,\n",
+       " (291, 159): 727,\n",
+       " (107, 115): 728,\n",
+       " (311, 302): 729,\n",
+       " (509, 112): 730,\n",
+       " (296, 263): 731,\n",
+       " (292, 168): 732,\n",
+       " (105, 98): 733,\n",
+       " (340, 115): 734,\n",
+       " (97, 118): 735,\n",
+       " (293, 566): 736,\n",
+       " (114, 117): 737,\n",
+       " (406, 319): 738,\n",
+       " (425, 610): 739,\n",
+       " (453, 116): 740,\n",
+       " (525, 323): 741,\n",
+       " (273, 373): 742,\n",
+       " (333, 115): 743,\n",
+       " (32, 106): 744,\n",
+       " (264, 174): 745,\n",
+       " (308, 121): 746,\n",
+       " (114, 121): 747,\n",
+       " (257, 163): 748,\n",
+       " (105, 122): 749,\n",
+       " (321, 274): 750,\n",
+       " (322, 412): 751,\n",
+       " (613, 431): 752,\n",
+       " (286, 108): 753,\n",
+       " (291, 183): 754,\n",
+       " (278, 111): 755,\n",
+       " (265, 100): 756,\n",
+       " (409, 114): 757,\n",
+       " (105, 101): 758,\n",
+       " (259, 175): 759,\n",
+       " (297, 109): 760,\n",
+       " (115, 111): 761,\n",
+       " (99, 99): 762,\n",
+       " (347, 108): 763,\n",
+       " (597, 313): 764,\n",
+       " (105, 97): 765,\n",
+       " (324, 115): 766,\n",
+       " (97, 396): 767,\n",
+       " (450, 275): 768,\n",
+       " (393, 101): 769,\n",
+       " (489, 274): 770,\n",
+       " (338, 100): 771,\n",
+       " (259, 150): 772,\n",
+       " (362, 116): 773,\n",
+       " (32, 51): 774,\n",
+       " (259, 173): 775,\n",
+       " (259, 154): 776,\n",
+       " (434, 116): 777,\n",
+       " (261, 137): 778,\n",
+       " (105, 270): 779,\n",
+       " (256, 167): 780,\n",
+       " (281, 161): 781,\n",
+       " (281, 184): 782,\n",
+       " (111, 111): 783,\n",
+       " (282, 276): 784,\n",
+       " (261, 175): 785,\n",
+       " (273, 662): 786,\n",
+       " (440, 111): 787,\n",
+       " (302, 378): 788,\n",
+       " (259, 171): 789,\n",
+       " (262, 101): 790,\n",
+       " (283, 101): 791,\n",
+       " (288, 378): 792,\n",
+       " (275, 354): 793,\n",
+       " (108, 371): 794,\n",
+       " (430, 521): 795,\n",
+       " (316, 290): 796,\n",
+       " (402, 115): 797,\n",
+       " (425, 294): 798,\n",
+       " (97, 378): 799,\n",
+       " (257, 182): 800,\n",
+       " (98, 276): 801,\n",
+       " (298, 184): 802,\n",
+       " (426, 101): 803,\n",
+       " (313, 460): 804,\n",
+       " (281, 181): 805,\n",
+       " (112, 568): 806,\n",
+       " (117, 306): 807,\n",
+       " (364, 56): 808,\n",
+       " (669, 49): 809,\n",
+       " (270, 282): 810,\n",
+       " (290, 104): 811,\n",
+       " (110, 371): 812,\n",
+       " (272, 109): 813,\n",
+       " (262, 370): 814,\n",
+       " (264, 170): 815,\n",
+       " (456, 761): 816,\n",
+       " (256, 173): 817,\n",
+       " (261, 182): 818,\n",
+       " (278, 97): 819,\n",
+       " (302, 103): 820,\n",
+       " (291, 184): 821,\n",
+       " (410, 121): 822,\n",
+       " (588, 149): 823,\n",
+       " (550, 325): 824,\n",
+       " (390, 110): 825,\n",
+       " (349, 330): 826,\n",
+       " (264, 149): 827,\n",
+       " (270, 101): 828,\n",
+       " (344, 326): 829,\n",
+       " (264, 165): 830,\n",
+       " (446, 274): 831,\n",
+       " (318, 101): 832,\n",
+       " (261, 154): 833,\n",
+       " (390, 112): 834,\n",
+       " (385, 116): 835,\n",
+       " (301, 276): 836,\n",
+       " (301, 270): 837,\n",
+       " (406, 425): 838,\n",
+       " (398, 110): 839,\n",
+       " (257, 171): 840,\n",
+       " (331, 298): 841,\n",
+       " (298, 166): 842,\n",
+       " (423, 489): 843,\n",
+       " (334, 118): 844,\n",
+       " (625, 644): 845,\n",
+       " (256, 154): 846,\n",
+       " (272, 270): 847,\n",
+       " (281, 174): 848,\n",
+       " (278, 121): 849,\n",
+       " (301, 594): 850,\n",
+       " (540, 107): 851,\n",
+       " (265, 114): 852,\n",
+       " (446, 408): 853,\n",
+       " (99, 289): 854,\n",
+       " (357, 288): 855,\n",
+       " (267, 116): 856,\n",
+       " (296, 294): 857,\n",
+       " (257, 167): 858,\n",
+       " (105, 501): 859,\n",
+       " (439, 292): 860,\n",
+       " (307, 285): 861,\n",
+       " (259, 161): 862,\n",
+       " (680, 290): 863,\n",
+       " (509, 109): 864,\n",
+       " (595, 115): 865,\n",
+       " (295, 101): 866,\n",
+       " (111, 365): 867,\n",
+       " (661, 110): 868,\n",
+       " (105, 323): 869,\n",
+       " (97, 117): 870,\n",
+       " (307, 547): 871,\n",
+       " (260, 140): 872,\n",
+       " (325, 269): 873,\n",
+       " (641, 111): 874,\n",
+       " (278, 99): 875,\n",
+       " (466, 309): 876,\n",
+       " (259, 159): 877,\n",
+       " (307, 616): 878,\n",
+       " (331, 455): 879,\n",
+       " (278, 483): 880,\n",
+       " (278, 266): 881,\n",
+       " (111, 468): 882,\n",
+       " (375, 100): 883,\n",
+       " (435, 432): 884,\n",
+       " (331, 392): 885,\n",
+       " (261, 178): 886,\n",
+       " (399, 584): 887,\n",
+       " (415, 115): 888,\n",
+       " (278, 623): 889,\n",
+       " (329, 302): 890,\n",
+       " (97, 338): 891,\n",
+       " (324, 101): 892,\n",
+       " (285, 276): 893,\n",
+       " (301, 101): 894,\n",
+       " (603, 313): 895,\n",
+       " (271, 181): 896,\n",
+       " (527, 346): 897,\n",
+       " (448, 102): 898,\n",
+       " (32, 89): 899,\n",
+       " (291, 178): 900,\n",
+       " (117, 101): 901,\n",
+       " (264, 154): 902,\n",
+       " (261, 161): 903,\n",
+       " (319, 263): 904,\n",
+       " (322, 371): 905,\n",
+       " (278, 112): 906,\n",
+       " (101, 119): 907,\n",
+       " (288, 115): 908,\n",
+       " (117, 98): 909,\n",
+       " (343, 111): 910,\n",
+       " (353, 288): 911,\n",
+       " (262, 119): 912,\n",
+       " (32, 75): 913,\n",
+       " (602, 812): 914,\n",
+       " (295, 326): 915,\n",
+       " (667, 369): 916,\n",
+       " (353, 99): 917,\n",
+       " (41, 46): 918,\n",
+       " (330, 539): 919,\n",
+       " (814, 101): 920,\n",
+       " (407, 702): 921,\n",
+       " (298, 130): 922,\n",
+       " (32, 52): 923,\n",
+       " (107, 305): 924,\n",
+       " (297, 121): 925,\n",
+       " (101, 112): 926,\n",
+       " (32, 448): 927,\n",
+       " (273, 410): 928,\n",
+       " (298, 135): 929,\n",
+       " (297, 100): 930,\n",
+       " (841, 165): 931,\n",
+       " (32, 397): 932,\n",
+       " (335, 294): 933,\n",
+       " (374, 320): 934,\n",
+       " (310, 102): 935,\n",
+       " (297, 115): 936,\n",
+       " (491, 884): 937,\n",
+       " (32, 561): 938,\n",
+       " (341, 781): 939,\n",
+       " (300, 405): 940,\n",
+       " (628, 677): 941,\n",
+       " (281, 149): 942,\n",
+       " (298, 170): 943,\n",
+       " (316, 898): 944,\n",
+       " (348, 477): 945,\n",
+       " (105, 302): 946,\n",
+       " (395, 104): 947,\n",
+       " (500, 323): 948,\n",
+       " (261, 136): 949,\n",
+       " (265, 103): 950,\n",
+       " (350, 363): 951,\n",
+       " (116, 101): 952,\n",
+       " (336, 859): 953,\n",
+       " (119, 110): 954,\n",
+       " (112, 323): 955,\n",
+       " (353, 116): 956,\n",
+       " (284, 592): 957,\n",
+       " (322, 370): 958,\n",
+       " (298, 156): 959,\n",
+       " (283, 100): 960,\n",
+       " (110, 305): 961,\n",
+       " (265, 578): 962,\n",
+       " (311, 415): 963,\n",
+       " (109, 276): 964,\n",
+       " (384, 296): 965,\n",
+       " (108, 108): 966,\n",
+       " (311, 121): 967,\n",
+       " (627, 616): 968,\n",
+       " (32, 328): 969,\n",
+       " (416, 320): 970,\n",
+       " (258, 140): 971,\n",
+       " (318, 107): 972,\n",
+       " (102, 449): 973,\n",
+       " (105, 685): 974,\n",
+       " (436, 121): 975,\n",
+       " (313, 294): 976,\n",
+       " (626, 604): 977,\n",
+       " (440, 282): 978,\n",
+       " (99, 270): 979,\n",
+       " (588, 183): 980,\n",
+       " (261, 159): 981,\n",
+       " (458, 432): 982,\n",
+       " (321, 656): 983,\n",
+       " (438, 115): 984,\n",
+       " (276, 118): 985,\n",
+       " (352, 274): 986,\n",
+       " (400, 277): 987,\n",
+       " (319, 274): 988,\n",
+       " (111, 347): 989,\n",
+       " (310, 100): 990,\n",
+       " (320, 503): 991,\n",
+       " (273, 403): 992,\n",
+       " (277, 315): 993,\n",
+       " (390, 365): 994,\n",
+       " (304, 102): 995,\n",
+       " (41, 44): 996,\n",
+       " (283, 115): 997,\n",
+       " (272, 365): 998,\n",
+       " (340, 99): 999,\n",
+       " (617, 463): 1000,\n",
+       " (291, 185): 1001,\n",
+       " (305, 115): 1002,\n",
+       " (32, 86): 1003,\n",
+       " (894, 948): 1004,\n",
+       " (270, 110): 1005,\n",
+       " (265, 344): 1006,\n",
+       " (256, 183): 1007,\n",
+       " (375, 634): 1008,\n",
+       " (114, 295): 1009,\n",
+       " (390, 893): 1010,\n",
+       " (461, 526): 1011,\n",
+       " (376, 115): 1012,\n",
+       " (281, 156): 1013,\n",
+       " (112, 104): 1014,\n",
+       " (119, 101): 1015,\n",
+       " (316, 289): 1016,\n",
+       " (271, 174): 1017,\n",
+       " (384, 632): 1018,\n",
+       " (912, 111): 1019,\n",
+       " (261, 173): 1020,\n",
+       " (32, 53): 1021,\n",
+       " (548, 350): 1022,\n",
+       " (306, 355): 1023,\n",
+       " (121, 115): 1024,\n",
+       " (108, 289): 1025,\n",
+       " (365, 108): 1026,\n",
+       " (262, 397): 1027,\n",
+       " (391, 184): 1028,\n",
+       " (409, 108): 1029,\n",
+       " (298, 175): 1030,\n",
+       " (290, 115): 1031,\n",
+       " (106, 590): 1032,\n",
+       " (849, 945): 1033,\n",
+       " (286, 111): 1034,\n",
+       " (109, 115): 1035,\n",
+       " (269, 414): 1036,\n",
+       " (571, 119): 1037,\n",
+       " (330, 461): 1038,\n",
+       " (323, 344): 1039,\n",
+       " (318, 116): 1040,\n",
+       " (473, 101): 1041,\n",
+       " (332, 754): 1042,\n",
+       " (340, 108): 1043,\n",
+       " (1029, 112): 1044,\n",
+       " (295, 537): 1045,\n",
+       " (870, 365): 1046,\n",
+       " (291, 149): 1047,\n",
+       " (308, 107): 1048,\n",
+       " (282, 289): 1049,\n",
+       " (393, 355): 1050,\n",
+       " (423, 720): 1051,\n",
+       " (294, 319): 1052,\n",
+       " (410, 115): 1053,\n",
+       " (311, 659): 1054,\n",
+       " (257, 173): 1055,\n",
+       " (257, 136): 1056,\n",
+       " (290, 101): 1057,\n",
+       " (717, 584): 1058,\n",
+       " (552, 771): 1059,\n",
+       " (315, 586): 1060,\n",
+       " (940, 348): 1061,\n",
+       " (453, 115): 1062,\n",
+       " (455, 161): 1063,\n",
+       " (262, 114): 1064,\n",
+       " (470, 104): 1065,\n",
+       " (657, 296): 1066,\n",
+       " (564, 592): 1067,\n",
+       " (434, 308): 1068,\n",
+       " (424, 579): 1069,\n",
+       " (429, 263): 1070,\n",
+       " (447, 271): 1071,\n",
+       " (783, 107): 1072,\n",
+       " (102, 434): 1073,\n",
+       " (660, 976): 1074,\n",
+       " (321, 420): 1075,\n",
+       " (306, 498): 1076,\n",
+       " (729, 121): 1077,\n",
+       " (823, 294): 1078,\n",
+       " (458, 325): 1079,\n",
+       " (298, 181): 1080,\n",
+       " (264, 172): 1081,\n",
+       " (299, 108): 1082,\n",
+       " (256, 150): 1083,\n",
+       " (124, 124): 1084,\n",
+       " (301, 111): 1085,\n",
+       " (296, 591): 1086,\n",
+       " (258, 131): 1087,\n",
+       " (265, 109): 1088,\n",
+       " (105, 387): 1089,\n",
+       " (281, 164): 1090,\n",
+       " (449, 110): 1091,\n",
+       " (266, 100): 1092,\n",
+       " (359, 441): 1093,\n",
+       " (354, 275): 1094,\n",
+       " (286, 375): 1095,\n",
+       " (332, 275): 1096,\n",
+       " (32, 54): 1097,\n",
+       " (257, 143): 1098,\n",
+       " (944, 621): 1099,\n",
+       " (300, 108): 1100,\n",
+       " (49, 57): 1101,\n",
+       " (350, 275): 1102,\n",
+       " (311, 468): 1103,\n",
+       " (416, 673): 1104,\n",
+       " (617, 296): 1105,\n",
+       " (281, 170): 1106,\n",
+       " (99, 303): 1107,\n",
+       " (880, 99): 1108,\n",
+       " (286, 592): 1109,\n",
+       " (448, 101): 1110,\n",
+       " (377, 486): 1111,\n",
+       " (99, 497): 1112,\n",
+       " (120, 116): 1113,\n",
+       " (763, 500): 1114,\n",
+       " (622, 107): 1115,\n",
+       " (260, 131): 1116,\n",
+       " (631, 563): 1117,\n",
+       " (282, 305): 1118,\n",
+       " (508, 396): 1119,\n",
+       " (475, 692): 1120,\n",
+       " (448, 318): 1121,\n",
+       " (474, 408): 1122,\n",
+       " (488, 986): 1123,\n",
+       " (300, 101): 1124,\n",
+       " (457, 269): 1125,\n",
+       " (32, 32): 1126,\n",
+       " (273, 98): 1127,\n",
+       " (102, 512): 1128,\n",
+       " (261, 171): 1129,\n",
+       " (824, 1001): 1130,\n",
+       " (325, 315): 1131,\n",
+       " (270, 100): 1132,\n",
+       " (695, 590): 1133,\n",
+       " (282, 533): 1134,\n",
+       " (423, 346): 1135,\n",
+       " (531, 275): 1136,\n",
+       " (265, 762): 1137,\n",
+       " (301, 114): 1138,\n",
+       " (329, 968): 1139,\n",
+       " (284, 598): 1140,\n",
+       " (579, 549): 1141,\n",
+       " (511, 182): 1142,\n",
+       " (278, 403): 1143,\n",
+       " (105, 303): 1144,\n",
+       " (357, 387): 1145,\n",
+       " (619, 165): 1146,\n",
+       " (266, 110): 1147,\n",
+       " (111, 118): 1148,\n",
+       " (276, 116): 1149,\n",
+       " (451, 723): 1150,\n",
+       " (265, 973): 1151,\n",
+       " (819, 372): 1152,\n",
+       " (451, 118): 1153,\n",
+       " (310, 1058): 1154,\n",
+       " (256, 171): 1155,\n",
+       " (291, 170): 1156,\n",
+       " (402, 306): 1157,\n",
+       " (532, 115): 1158,\n",
+       " (556, 115): 1159,\n",
+       " (369, 275): 1160,\n",
+       " (301, 104): 1161,\n",
+       " (531, 315): 1162,\n",
+       " (284, 784): 1163,\n",
+       " (272, 110): 1164,\n",
+       " (412, 101): 1165,\n",
+       " (317, 274): 1166,\n",
+       " (340, 109): 1167,\n",
+       " (117, 99): 1168,\n",
+       " (371, 110): 1169,\n",
+       " (403, 110): 1170,\n",
+       " (264, 161): 1171,\n",
+       " (384, 319): 1172,\n",
+       " (116, 305): 1173,\n",
+       " (366, 104): 1174,\n",
+       " (98, 323): 1175,\n",
+       " (646, 303): 1176,\n",
+       " (282, 104): 1177,\n",
+       " (671, 346): 1178,\n",
+       " (298, 136): 1179,\n",
+       " (306, 115): 1180,\n",
+       " (482, 115): 1181,\n",
+       " (422, 115): 1182,\n",
+       " (308, 110): 1183,\n",
+       " (627, 285): 1184,\n",
+       " (555, 770): 1185,\n",
+       " (108, 318): 1186,\n",
+       " (545, 101): 1187,\n",
+       " (416, 718): 1188,\n",
+       " (257, 137): 1189,\n",
+       " (340, 103): 1190,\n",
+       " (105, 333): 1191,\n",
+       " (267, 107): 1192,\n",
+       " (921, 115): 1193,\n",
+       " (99, 121): 1194,\n",
+       " (1015, 288): 1195,\n",
+       " (116, 323): 1196,\n",
+       " (529, 1114): 1197,\n",
+       " (838, 463): 1198,\n",
+       " (308, 115): 1199,\n",
+       " (424, 1066): 1200,\n",
+       " (267, 289): 1201,\n",
+       " (406, 549): 1202,\n",
+       " (674, 110): 1203,\n",
+       " (401, 417): 1204,\n",
+       " (391, 154): 1205,\n",
+       " (292, 130): 1206,\n",
+       " (1120, 1125): 1207,\n",
+       " (560, 115): 1208,\n",
+       " (281, 163): 1209,\n",
+       " (101, 285): 1210,\n",
+       " (505, 98): 1211,\n",
+       " (318, 115): 1212,\n",
+       " (278, 109): 1213,\n",
+       " (508, 98): 1214,\n",
+       " (291, 174): 1215,\n",
+       " (349, 479): 1216,\n",
+       " (571, 303): 1217,\n",
+       " (775, 315): 1218,\n",
+       " (266, 115): 1219,\n",
+       " (110, 101): 1220,\n",
+       " (103, 302): 1221,\n",
+       " (508, 578): 1222,\n",
+       " (117, 396): 1223,\n",
+       " (403, 305): 1224,\n",
+       " (321, 294): 1225,\n",
+       " (491, 959): 1226,\n",
+       " (111, 98): 1227,\n",
+       " (895, 274): 1228,\n",
+       " (1205, 902): 1229,\n",
+       " (349, 830): 1230,\n",
+       " (661, 270): 1231,\n",
+       " (668, 112): 1232,\n",
+       " (308, 396): 1233,\n",
+       " (264, 167): 1234,\n",
+       " (274, 296): 1235,\n",
+       " (621, 100): 1236,\n",
+       " (265, 99): 1237,\n",
+       " (439, 414): 1238,\n",
+       " (605, 467): 1239,\n",
+       " (366, 465): 1240,\n",
+       " (507, 147): 1241,\n",
+       " (956, 1195): 1242,\n",
+       " (302, 614): 1243,\n",
+       " (116, 288): 1244,\n",
+       " (518, 116): 1245,\n",
+       " (697, 263): 1246,\n",
+       " (341, 486): 1247,\n",
+       " (320, 354): 1248,\n",
+       " (331, 315): 1249,\n",
+       " (733, 323): 1250,\n",
+       " (843, 719): 1251,\n",
+       " (276, 109): 1252,\n",
+       " (446, 460): 1253,\n",
+       " (310, 118): 1254,\n",
+       " (534, 57): 1255,\n",
+       " ...}"
+      ]
+     },
+     "execution_count": 10,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "merges"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def _encode_chunk(chunk_bytes: bytes, verbose=False) -> list[int]:\n",
+    "    tokens = list(chunk_bytes)\n",
+    "    while len(tokens) >= 2:\n",
+    "        if verbose:\n",
+    "            visualise_tokens([vocab[token] for token in tokens]) # token can be > 256 after merging\n",
+    "        stats = {}\n",
+    "        get_stats(tokens, stats)\n",
+    "        pair = min(stats, key=lambda p: merges.get(p, float(\"inf\")))\n",
+    "        if not pair in merges:\n",
+    "            break\n",
+    "        idx = merges[pair]\n",
+    "        tokens = merge(tokens, pair, idx)\n",
+    "    return tokens\n",
+    "\n",
+    "def encode_ordinary(text, verbose=False) -> list[int]:\n",
+    "    chunk_texts = re.findall(regex, text)\n",
+    "    ids_list = []\n",
+    "    for i, text in enumerate(chunk_texts):\n",
+    "        if verbose:\n",
+    "            print()\n",
+    "            print(f\"encoding chunk {i+1}/{len(chunk_texts)}: {text}\")\n",
+    "        chunk_bytes = text.encode(\"utf-8\") # raw bytes\n",
+    "        ids = _encode_chunk(chunk_bytes, verbose)\n",
+    "        ids_list.extend(ids)\n",
+    "    return ids_list\n",
+    "\n",
+    "def encode(text, verbose=False, allowed_special=\"none\") -> list[int]:\n",
+    "    special = {}\n",
+    "    if allowed_special == \"all\":\n",
+    "        special = special_tokens\n",
+    "    elif allowed_special == \"none\":\n",
+    "        special = {}\n",
+    "    elif allowed_special == \"none_raise\":\n",
+    "        special = {}\n",
+    "        assert all(token not in text for token in special_tokens), \"Text contains special tokens that are not allowed\"\n",
+    "    elif isinstance(allowed_special, set):\n",
+    "        special = {k: v for k, v in special_tokens.items() if k in allowed_special}\n",
+    "    else:\n",
+    "        raise ValueError(f\"allowed_special={allowed_special} not understood.\")\n",
+    "    if not special:\n",
+    "        return encode_ordinary(text, verbose)\n",
+    "    special_pattern = \"(\" + \"|\".join(re.escape(token) for token in special) + \")\"\n",
+    "    parts = re.split(special_pattern, text)\n",
+    "    ids = []\n",
+    "    for part in parts:\n",
+    "        if part in special:\n",
+    "            ids.append(special[part])\n",
+    "        else:\n",
+    "            ids.extend(encode_ordinary(part, verbose))\n",
+    "    return ids"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।\n",
+      "[2637, 665, 666, 320, 443, 441, 2459, 531, 332, 752, 374, 435, 257, 129, 44, 1919, 2649, 462, 1462, 274, 389]\n",
+      "\n",
+      "मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together.\n",
+      "[392, 872, 369, 392, 293, 577, 873, 458, 725, 291, 155, 269, 466, 33, 1268, 321, 486, 2662, 933, 462, 383, 268, 44, 2127, 39, 115, 678, 106, 1302, 2510, 46]\n",
+      "\n",
+      "स्वल्पा adjust करो, बैंगलोर का ट्रैफिक ऐसा ही है।\n",
+      "[369, 692, 354, 1156, 269, 756, 106, 648, 631, 320, 44, 443, 379, 1538, 354, 1751, 620, 877, 2627, 840, 510, 259, 144, 369, 269, 2581, 466, 389]\n",
+      "\n",
+      "ನೀವು ಚಹಾ ಕುಡಿತೀರಾ? मुझे एक cup चाहिए।\n",
+      "[317, 464, 750, 833, 596, 279, 2112, 381, 488, 464, 296, 279, 63, 345, 580, 157, 275, 583, 286, 1989, 776, 1378, 642, 389]\n",
+      "\n",
+      "आज का काम पूरा करो, ನಾಳೆ ಎಲ್ಲಿಂದ ಆರಂಭಿಸೋದು ನೋಡಿ।\n",
+      "[2637, 665, 620, 3169, 3092, 277, 269, 631, 320, 44, 424, 2810, 294, 555, 418, 764, 1986, 1529, 406, 428, 3004, 424, 2826, 268, 389]\n",
+      "\n",
+      "ಪಾರ್ಟಿ ಹೇಗೆ ಇತ್ತು? मुझे तो बहुत मजा आया!\n",
+      "[515, 417, 526, 268, 452, 384, 933, 603, 472, 63, 345, 580, 157, 275, 666, 320, 443, 441, 2459, 345, 665, 269, 640, 257, 431, 33]\n",
+      "\n",
+      "ನಮ್ಮ ಚೂರು ಸಹನಶೀಲತೆಯನ್ನು ತೋರಿಸಿ, ये थोड़ी देर का मसला है।\n",
+      "[317, 1552, 833, 486, 463, 1389, 317, 632, 464, 319, 327, 1122, 618, 1747, 2695, 44, 759, 275, 531, 320, 977, 315, 475, 1452, 620, 345, 369, 1515, 466, 389]\n",
+      "\n",
+      "ಸಮಯ ನಿಲ್ಲುತ್ತಿಲ್ಲ, जिंदगी में स्वल्पा मज़ा भी जरूरी है।\n",
+      "[429, 479, 574, 424, 1289, 660, 1289, 44, 416, 1206, 633, 503, 315, 404, 1508, 354, 1156, 269, 345, 1440, 269, 1218, 416, 277, 435, 993, 466, 389]\n",
+      "\n",
+      "My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>\n",
+      "[77, 121, 2142, 356, 676, 101, 695, 1717, 122, 556, 44, 312, 362, 39, 109, 272, 1323, 276, 304, 376, 109, 2730, 283, 46, 3256]\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "for test in tests:\n",
+    "    print(test)\n",
+    "    print(encode(test, allowed_special={\"<|endoftext|>\"}))\n",
+    "    print()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Text: आज तो बहुत थक गया हूँ, ಸ್ವಲ್ಪ विश्रಾಂತಿ ಬೇಕು।\n",
+      "Tokens: [2637, 665, 666, 320, 443, 441, 2459, 531, 332, 752, 374, 435, 257, 129, 44, 1919, 2649, 462, 1462, 274, 389]\n",
+      "\n",
+      "Unmerged length: 117\n",
+      "Merged length: 21\n",
+      "--------------------------------------------------\n",
+      "Text: मौसम कितना अच्छा है! ನೀವೂ ಹೊರಗೆ ಬನ್ನಿ, let's enjoy together.\n",
+      "Tokens: [392, 872, 369, 392, 293, 577, 873, 458, 725, 291, 155, 269, 466, 33, 1268, 321, 486, 2662, 933, 462, 383, 268, 44, 2127, 39, 115, 678, 106, 1302, 2510, 46]\n",
+      "\n",
+      "Unmerged length: 120\n",
+      "Merged length: 31\n",
+      "--------------------------------------------------\n",
+      "Text: स्वल्पा adjust करो, बैंगलोर का ट्रैफिक ऐसा ही है।\n",
+      "Tokens: [369, 692, 354, 1156, 269, 756, 106, 648, 631, 320, 44, 443, 379, 1538, 354, 1751, 620, 877, 2627, 840, 510, 259, 144, 369, 269, 2581, 466, 389]\n",
+      "\n",
+      "Unmerged length: 117\n",
+      "Merged length: 28\n",
+      "--------------------------------------------------\n",
+      "Text: ನೀವು ಚಹಾ ಕುಡಿತೀರಾ? मुझे एक cup चाहिए।\n",
+      "Tokens: [317, 464, 750, 833, 596, 279, 2112, 381, 488, 464, 296, 279, 63, 345, 580, 157, 275, 583, 286, 1989, 776, 1378, 642, 389]\n",
+      "\n",
+      "Unmerged length: 91\n",
+      "Merged length: 24\n",
+      "--------------------------------------------------\n",
+      "Text: आज का काम पूरा करो, ನಾಳೆ ಎಲ್ಲಿಂದ ಆರಂಭಿಸೋದು ನೋಡಿ।\n",
+      "Tokens: [2637, 665, 620, 3169, 3092, 277, 269, 631, 320, 44, 424, 2810, 294, 555, 418, 764, 1986, 1529, 406, 428, 3004, 424, 2826, 268, 389]\n",
+      "\n",
+      "Unmerged length: 126\n",
+      "Merged length: 25\n",
+      "--------------------------------------------------\n",
+      "Text: ಪಾರ್ಟಿ ಹೇಗೆ ಇತ್ತು? मुझे तो बहुत मजा आया!\n",
+      "Tokens: [515, 417, 526, 268, 452, 384, 933, 603, 472, 63, 345, 580, 157, 275, 666, 320, 443, 441, 2459, 345, 665, 269, 640, 257, 431, 33]\n",
+      "\n",
+      "Unmerged length: 102\n",
+      "Merged length: 26\n",
+      "--------------------------------------------------\n",
+      "Text: ನಮ್ಮ ಚೂರು ಸಹನಶೀಲತೆಯನ್ನು ತೋರಿಸಿ, ये थोड़ी देर का मसला है।\n",
+      "Tokens: [317, 1552, 833, 486, 463, 1389, 317, 632, 464, 319, 327, 1122, 618, 1747, 2695, 44, 759, 275, 531, 320, 977, 315, 475, 1452, 620, 345, 369, 1515, 466, 389]\n",
+      "\n",
+      "Unmerged length: 148\n",
+      "Merged length: 30\n",
+      "--------------------------------------------------\n",
+      "Text: ಸಮಯ ನಿಲ್ಲುತ್ತಿಲ್ಲ, जिंदगी में स्वल्पा मज़ा भी जरूरी है।\n",
+      "Tokens: [429, 479, 574, 424, 1289, 660, 1289, 44, 416, 1206, 633, 503, 315, 404, 1508, 354, 1156, 269, 345, 1440, 269, 1218, 416, 277, 435, 993, 466, 389]\n",
+      "\n",
+      "Unmerged length: 147\n",
+      "Merged length: 28\n",
+      "--------------------------------------------------\n",
+      "Text: My name is Jeff Bezos, and I'm the owner of Amazon.<|endoftext|>\n",
+      "Tokens: [77, 121, 2142, 356, 676, 101, 695, 1717, 122, 556, 44, 312, 362, 39, 109, 272, 1323, 276, 304, 376, 109, 2730, 283, 46, 3256]\n",
+      "\n",
+      "Unmerged length: 64\n",
+      "Merged length: 25\n",
+      "--------------------------------------------------\n"
+     ]
+    }
+   ],
+   "source": [
+    "def test_encoding(text, verbose=False, allowed_special=None):\n",
+    "    print(f\"Text: {text}\")\n",
+    "    test_ids = encode(text, verbose=verbose, allowed_special=allowed_special)\n",
+    "    print(f\"Tokens: {test_ids}\")\n",
+    "    print(\"\")\n",
+    "    print(f\"Unmerged length: {len(text.encode('utf-8'))}\")\n",
+    "    print(f\"Merged length: {len(test_ids)}\")\n",
+    "    print(\"-\"*50)\n",
+    "\n",
+    "for test in tests:\n",
+    "    test_encoding(test, allowed_special=\"all\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "virtual",
+   "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": "3.11.7"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

bpe_tok.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c6d964893049f0baaa2bb920e3c71b85ff43cad3c5b3a51352fb9177764d26b
+size 24563

bpe_tok.vocab

@@ -0,0 +1,3256 @@
+[\u0000] 0
+[\u0001] 1
+[\u0002] 2
+[\u0003] 3
+[\u0004] 4
+[\u0005] 5
+[\u0006] 6
+[\u0007] 7
+[\u0008] 8
+[\u0009] 9
+[\u000a] 10
+[\u000b] 11
+[\u000c] 12
+[\u000d] 13
+[\u000e] 14
+[\u000f] 15
+[\u0010] 16
+[\u0011] 17
+[\u0012] 18
+[\u0013] 19
+[\u0014] 20
+[\u0015] 21
+[\u0016] 22
+[\u0017] 23
+[\u0018] 24
+[\u0019] 25
+[\u001a] 26
+[\u001b] 27
+[\u001c] 28
+[\u001d] 29
+[\u001e] 30
+[\u001f] 31
+[ ] 32
+[!] 33
+["] 34
+[#] 35
+[$] 36
+[%] 37
+[&] 38
+['] 39
+[(] 40
+[)] 41
+[*] 42
+[+] 43
+[,] 44
+[-] 45
+[.] 46
+[/] 47
+[0] 48
+[1] 49
+[2] 50
+[3] 51
+[4] 52
+[5] 53
+[6] 54
+[7] 55
+[8] 56
+[9] 57
+[:] 58
+[;] 59
+[<] 60
+[=] 61
+[>] 62
+[?] 63
+[@] 64
+[A] 65
+[B] 66
+[C] 67
+[D] 68
+[E] 69
+[F] 70
+[G] 71
+[H] 72
+[I] 73
+[J] 74
+[K] 75
+[L] 76
+[M] 77
+[N] 78
+[O] 79
+[P] 80
+[Q] 81
+[R] 82
+[S] 83
+[T] 84
+[U] 85
+[V] 86
+[W] 87
+[X] 88
+[Y] 89
+[Z] 90
+[[] 91
+[\] 92
+[]] 93
+[^] 94
+[_] 95
+[`] 96
+[a] 97
+[b] 98
+[c] 99
+[d] 100
+[e] 101
+[f] 102
+[g] 103
+[h] 104
+[i] 105
+[j] 106
+[k] 107
+[l] 108
+[m] 109
+[n] 110
+[o] 111
+[p] 112
+[q] 113
+[r] 114
+[s] 115
+[t] 116
+[u] 117
+[v] 118
+[w] 119
+[x] 120
+[y] 121
+[z] 122
+[{] 123
+[|] 124
+[}] 125
+[~] 126
+[\u007f] 127
+[�] 128
+[�] 129
+[�] 130
+[�] 131
+[�] 132
+[�] 133
+[�] 134
+[�] 135
+[�] 136
+[�] 137
+[�] 138
+[�] 139
+[�] 140
+[�] 141
+[�] 142
+[�] 143
+[�] 144
+[�] 145
+[�] 146
+[�] 147
+[�] 148
+[�] 149
+[�] 150
+[�] 151
+[�] 152
+[�] 153
+[�] 154
+[�] 155
+[�] 156
+[�] 157
+[�] 158
+[�] 159
+[�] 160
+[�] 161
+[�] 162
+[�] 163
+[�] 164
+[�] 165
+[�] 166
+[�] 167
+[�] 168
+[�] 169
+[�] 170
+[�] 171
+[�] 172
+[�] 173
+[�] 174
+[�] 175
+[�] 176
+[�] 177
+[�] 178
+[�] 179
+[�] 180
+[�] 181
+[�] 182
+[�] 183
+[�] 184
+[�] 185
+[�] 186
+[�] 187
+[�] 188
+[�] 189
+[�] 190
+[�] 191
+[�] 192
+[�] 193
+[�] 194
+[�] 195
+[�] 196
+[�] 197
+[�] 198
+[�] 199
+[�] 200
+[�] 201
+[�] 202
+[�] 203
+[�] 204
+[�] 205
+[�] 206
+[�] 207
+[�] 208
+[�] 209
+[�] 210
+[�] 211
+[�] 212
+[�] 213
+[�] 214
+[�] 215
+[�] 216
+[�] 217
+[�] 218
+[�] 219
+[�] 220
+[�] 221
+[�] 222
+[�] 223
+[�] 224
+[�] 225
+[�] 226
+[�] 227
+[�] 228
+[�] 229
+[�] 230
+[�] 231
+[�] 232
+[�] 233
+[�] 234
+[�] 235
+[�] 236
+[�] 237
+[�] 238
+[�] 239
+[�] 240
+[�] 241
+[�] 242
+[�] 243
+[�] 244
+[�] 245
+[�] 246
+[�] 247
+[�] 248
+[�] 249
+[�] 250
+[�] 251
+[�] 252
+[�] 253
+[�] 254
+[�] 255
+[�][�] -> [�] 256
+[�][�] -> [�] 257
+[�][�] -> [�] 258
+[ ][�] -> [ �] 259
+[�][�] -> [�] 260
+[ ][�] -> [ �] 261
+[ ][t] -> [ t] 262
+[�][�] -> [್] 263
+[್][�] -> [್�] 264
+[ ][a] -> [ a] 265
+[h][e] -> [he] 266
+[i][n] -> [in] 267
+[�][�] -> [ಿ] 268
+[�][�] -> [ा] 269
+[r][e] -> [re] 270
+[ಿ][�] -> [ಿ�] 271
+[ t][he] -> [ the] 272
+[ ][o] -> [ o] 273
+[�][�] -> [ು] 274
+[�][�] -> [े] 275
+[e][r] -> [er] 276
+[�][�] -> [र] 277
+[ ][s] -> [ s] 278
+[�][�] -> [ಾ] 279
+[�][�] -> [्] 280
+[ಾ][�] -> [ಾ�] 281
+[a][t] -> [at] 282
+[o][n] -> [on] 283
+[ ][w] -> [ w] 284
+[n][d] -> [nd] 285
+[ ][c] -> [ c] 286
+[�][�] -> [ि] 287
+[e][n] -> [en] 288
+[e][s] -> [es] 289
+[i][s] -> [is] 290
+[्][�] -> [्�] 291
+[ि][�] -> [ि�] 292
+[ �][�] -> [ क] 293
+[�][�] -> [ೆ] 294
+[i][t] -> [it] 295
+[�][�] -> [ರ] 296
+[o][r] -> [or] 297
+[ा][�] -> [ा�] 298
+[ ][b] -> [ b] 299
+[ ][f] -> [ f] 300
+[ ][p] -> [ p] 301
+[a][n] -> [an] 302
+[e][d] -> [ed] 303
+[ o][f] -> [ of] 304
+[in][g] -> [ing] 305
+[a][l] -> [al] 306
+[o][u] -> [ou] 307
+[a][r] -> [ar] 308
+[�][�] -> [ं] 309
+[ ][in] -> [ in] 310
+[ ][m] -> [ m] 311
+[ a][nd] -> [ and] 312
+[�][�] -> [ದ] 313
+[ t][o] -> [ to] 314
+[�][�] -> [ी] 315
+[ ][d] -> [ d] 316
+[�][�] -> [ನ] 317
+[i][c] -> [ic] 318
+[�][�] -> [ಲ] 319
+[�][�] -> [ो] 320
+[�][�] -> [ವ] 321
+[ ][h] -> [ h] 322
+[l][e] -> [le] 323
+[a][s] -> [as] 324
+[�][�] -> [न] 325
+[i][on] -> [ion] 326
+[�][�] -> [ತ] 327
+[r][o] -> [ro] 328
+[ t][h] -> [ th] 329
+[�][�] -> [ಂ] 330
+[ �][�] -> [ स] 331
+[�][�] -> [क] 332
+[en][t] -> [ent] 333
+[ ][e] -> [ e] 334
+[�][�] -> [ಗ] 335
+[ ][l] -> [ l] 336
+[�][�] -> [�] 337
+[i][l] -> [il] 338
+[ ][T] -> [ T] 339
+[ ][re] -> [ re] 340
+[ �][�] -> [ ಮ] 341
+[o][m] -> [om] 342
+[ ][n] -> [ n] 343
+[c][t] -> [ct] 344
+[ �][�] -> [ म] 345
+[�][�] -> [ಳ] 346
+[v][e] -> [ve] 347
+[s][t] -> [st] 348
+[ �][�] -> [ ಸ] 349
+[�][�] -> [त] 350
+[್�][�] -> [್ನ] 351
+[್�][�] -> [್ತ] 352
+[ b][e] -> [ be] 353
+[�][�] -> [ल] 354
+[l][y] -> [ly] 355
+[ ][is] -> [ is] 356
+[ ][g] -> [ g] 357
+[�][�] -> [ಕ] 358
+[ �][�] -> [ प] 359
+[್�][�] -> [್ರ] 360
+[ क][े] -> [ के] 361
+[ ][I] -> [ I] 362
+[्][र] -> [���र] 363
+[ ][1] -> [ 1] 364
+[s][e] -> [se] 365
+[o][t] -> [ot] 366
+[े][ं] -> [ें] 367
+[್�][�] -> [್ಲ] 368
+[�][�] -> [स] 369
+[i][m] -> [im] 370
+[o][w] -> [ow] 371
+[i][d] -> [id] 372
+[u][t] -> [ut] 373
+[ �][�] -> [ ह] 374
+[o][l] -> [ol] 375
+[ ][A] -> [ A] 376
+[್�][�] -> [್ಯ] 377
+[c][e] -> [ce] 378
+[�][�] -> [ै] 379
+[ �][�] -> [ ಅ] 380
+[�][�] -> [ಡ] 381
+[ th][at] -> [ that] 382
+[ನ][್ನ] -> [ನ್ನ] 383
+[�][�] -> [ೇ] 384
+[ ][S] -> [ S] 385
+[ತ][್ತ] -> [ತ್ತ] 386
+[e][t] -> [et] 387
+[ �][�] -> [ ಪ] 388
+[�][�] -> [।] 389
+[ ][u] -> [ u] 390
+[ೆ][�] -> [ೆ�] 391
+[�][�] -> [म] 392
+[ o][n] -> [ on] 393
+[ f][or] -> [ for] 394
+[ ][C] -> [ C] 395
+[c][h] -> [ch] 396
+[r][a] -> [ra] 397
+[i][g] -> [ig] 398
+[ s][t] -> [ st] 399
+[�][�] -> [ु] 400
+[ �][�] -> [ ಕ] 401
+[at][ion] -> [ation] 402
+[u][r] -> [ur] 403
+[ म][ें] -> [ में] 404
+[i][r] -> [ir] 405
+[ಿ�][�] -> [ಿಸ] 406
+[ ][y] -> [ y] 407
+[ನ್ನ][ು] -> [ನ್ನು] 408
+[ ][he] -> [ he] 409
+[v][er] -> [ver] 410
+[ T][he] -> [ The] 411
+[a][d] -> [ad] 412
+[ ][(] -> [ (] 413
+[र][्�] -> [र्�] 414
+[a][y] -> [ay] 415
+[ �][�] -> [ ज] 416
+[ಾ�][�] -> [ಾರ] 417
+[ಲ][್ಲ] -> [ಲ್ಲ] 418
+[e][l] -> [el] 419
+[ಾ�][�] -> [ಾಗ] 420
+[ಿ�][�] -> [ಿಯ] 421
+[ a][s] -> [ as] 422
+[�][�] -> [ೊ] 423
+[ �][�] -> [ ನ] 424
+[ಿ�][�] -> [ಿದ] 425
+[a][m] -> [am] 426
+[it][h] -> [ith] 427
+[�][�] -> [ೋ] 428
+[�][�] -> [ಸ] 429
+[ ][2] -> [ 2] 430
+[�][ा] -> [�ा] 431
+[�][�] -> [प] 432
+[ a][re] -> [ are] 433
+[u][l] -> [ul] 434
+[�][�] -> [ू] 435
+[ a][n] -> [ an] 436
+[ w][ith] -> [ with] 437
+[ ][it] -> [ it] 438
+[ �][�] -> [ व] 439
+[ w][h] -> [ wh] 440
+[�][�] -> [ह] 441
+[ ][M] -> [ M] 442
+[ �][�] -> [ ब] 443
+[ y][ou] -> [ you] 444
+[ �][�] -> [ ए] 445
+[ಗ][ಳ] -> [ಗಳ] 446
+[ �][�] -> [ ವ] 447
+[i][f] -> [if] 448
+[t][er] -> [ter] 449
+[ �][�] -> [ न] 450
+[ p][ro] -> [ pro] 451
+[ �][�] -> [ ಹ] 452
+[ c][on] -> [ con] 453
+[ ][P] -> [ P] 454
+[ं][�] -> [ं�] 455
+[ a][l] -> [ al] 456
+[ा][र] -> [ार] 457
+[ �][�] -> [ अ] 458
+[ o][r] -> [ or] 459
+[ಲ್ಲ][ಿ] -> [ಲ್ಲಿ] 460
+[�][�] -> [ಟ] 461
+[ �][�] -> [ ಬ] 462
+[ರ][ು] -> [ರು] 463
+[�][�] -> [ೀ] 464
+[he][r] -> [her] 465
+[ ह][ै] -> [ है] 466
+[ो][ं] -> [ों] 467
+[o][d] -> [od] 468
+[er][s] -> [ers] 469
+[ ][W] -> [ W] 470
+[ ][B] -> [ B] 471
+[ತ್ತ][ು] -> [ತ್ತು] 472
+[ ][H] -> [ H] 473
+[ೆ�][�] -> [ೆಯ] 474
+[ �][�] -> [ द] 475
+[ �][�] -> [ औ] 476
+[e][m] -> [em] 477
+[ �][�] -> [ ल] 478
+[�][�] -> [ಮ] 479
+[�][�] -> [’] 480
+[ w][e] -> [ we] 481
+[a][nd] -> [and] 482
+[p][e] -> [pe] 483
+[ ಮ][ತ್ತು] -> [ ಮತ್ತು] 484
+[ರ][್�] -> [ರ್�] 485
+[�][�] -> [ೂ] 486
+[ औ][र] -> [ और] 487
+[ಿ�][�] -> [ಿತ] 488
+[ಂ][ದ] -> [ಂದ] 489
+[a][g] -> [ag] 490
+[ �][�] -> [ र] 491
+[ w][as] -> [ was] 492
+[ �][�] -> [ ग] 493
+[ e][x] -> [ ex] 494
+[l][d] -> [ld] 495
+[ಿ�][�] -> [ಿಕ] 496
+[es][s] -> [ess] 497
+[t][h] -> [th] 498
+[ �][�] -> [ ಆ] 499
+[o][p] -> [op] 500
+[k][e] -> [ke] 501
+[il][l] -> [ill] 502
+[�][�] -> [ग] 503
+[ ][E] -> [ E] 504
+[r][i] -> [ri] 505
+[्�][�] -> [्य] 506
+[ ][�] -> [ �] 507
+[ s][u] -> [ su] 508
+[ c][om] -> [ com] 509
+[ि�][�] -> [िक] 510
+[े][�] -> [े�] 511
+[o][re] -> [ore] 512
+[is][t] -> [ist] 513
+[ ][v] -> [ v] 514
+[�][�] -> [ಪ] 515
+[ h][a] -> [ ha] 516
+[es][t] -> [est] 517
+[ a][t] -> [ at] 518
+[u][m] -> [um] 519
+[ा�][�] -> [ान] 520
+[0][0] -> [00] 521
+[ b][y] -> [ by] 522
+[ ]["] -> [ "] 523
+[at][e] -> [ate] 524
+[a][b] -> [ab] 525
+[್�][�] -> [್ಟ] 526
+[ಿ�][�] -> [ಿಗ] 527
+[u][s] -> [us] 528
+[ d][e] -> [ de] 529
+[a][in] -> [ain] 530
+[ �][�] -> [ थ] 531
+[u][re] -> [ure] 532
+[i][ve] -> [ive] 533
+[ 1][9] -> [ 19] 534
+[u][n] -> [un] 535
+[ ][D] -> [ D] 536
+[i][es] -> [ies] 537
+[ig][h] -> [igh] 538
+[�][�] -> [ಬ] 539
+[a][c] -> [ac] 540
+[ro][m] -> [rom] 541
+[an][t] -> [ant] 542
+[ s][e] -> [ se] 543
+[ ][L] -> [ L] 544
+[ ][R] -> [ R] 545
+[�][�] -> [व] 546
+[n][t] -> [nt] 547
+[ प][्र] -> [ प्र] 548
+[ು][ವ] -> [ುವ] 549
+[ �][�] -> [ उ] 550
+[ क][ो] -> [ को] 551
+[ c][h] -> [ ch] 552
+[it][y] -> [ity] 553
+[ c][an] -> [ can] 554
+[ �][�] -> [ ಎ] 555
+[o][s] -> [os] 556
+[ n][ot] -> [ not] 557
+[ �][�] -> [ ರ] 558
+[ ಪ][್ರ] -> [ ಪ್ರ] 559
+[m][ent] -> [ment] 560
+[q][u] -> [qu] 561
+[ ][F] -> [ F] 562
+[न][े] -> [ने] 563
+[ s][h] -> [ sh] 564
+[ ][N] -> [ N] 565
+[ि�][�ा] -> [िया] 566
+[i][v] -> [iv] 567
+[or][t] -> [ort] 568
+[ स][े] -> [ से] 569
+[ f][rom] -> [ from] 570
+[ n][e] -> [ ne] 571
+[ಾ�][�] -> [ಾನ] 572
+[ಿ�][�] -> [ಿನ] 573
+[�][�] -> [ಯ] 574
+[ क][ी] -> [ की] 575
+[o][c] -> [oc] 576
+[ि�][�] -> [ित] 577
+[p][p] -> [pp] 578
+[ಿ�][�] -> [ಿರ] 579
+[ु][�] -> [ु�] 580
+[�][�] -> [ೈ] 581
+[ ha][ve] -> [ have] 582
+[ ए][क] -> [ एक] 583
+[u][d] -> [ud] 584
+[ ][r] -> [ r] 585
+[�][�] -> [य] 586
+[ ][G] -> [ G] 587
+[ಕ][್�] -> [ಕ್�] 588
+[�][�] -> [ब] 589
+[e][ct] -> [ect] 590
+[�][�] -> [ಣ] 591
+[ou][ld] -> [ould] 592
+[ ][-] -> [ -] 593
+[ar][t] -> [art] 594
+[re][s] -> [res] 595
+[�][�] -> [ಹ] 596
+[ಿ�][�] -> [ಿಂ] 597
+[el][l] -> [ell] 598
+[ ][le] -> [ le] 599
+[igh][t] -> [ight] 600
+[ನ][್] -> [ನ್] 601
+[ ][k] -> [ k] 602
+[ �][�] -> [ ಇ] 603
+[�][�] -> [़] 604
+[ि�][�] -> [िय] 605
+[ th][is] -> [ this] 606
+[ic][al] -> [ical] 607
+[i][al] -> [ial] 608
+[�][�] -> [”] 609
+[್�][�] -> [್ದ] 610
+[�][�] -> [ट] 611
+[ �][�] -> [ “] 612
+[ ग][�] -> [ ग�] 613
+[g][e] -> [ge] 614
+[ಾ�][�] -> [ಾದ] 615
+[g][h] -> [gh] 616
+[ ಅ][ವ] -> [ ಅವ] 617
+[ �][�] -> [ ತ] 618
+[ स][्�] -> [ स्�] 619
+[ क][ा] -> [ का] 620
+[e][re] -> [ere] 621
+[ w][or] -> [ wor] 622
+[om][e] -> [ome] 623
+[ic][h] -> [ich] 624
+[ a][b] -> [ ab] 625
+[�][�] -> [ड] 626
+[r][ou] -> [rou] 627
+[ಾ�][�] -> [ಾಯ] 628
+[ ][O] -> [ O] 629
+[in][e] -> [ine] 630
+[ क][र] -> [ कर] 631
+[�][�] -> [ಶ] 632
+[�][�] -> [द] 633
+[o][g] -> [og] 634
+[a][p] -> [ap] 635
+[al][l] -> [all] 636
+[�][�] -> [ॉ] 637
+[ar][d] -> [ard] 638
+[ा�][�] -> [ाल] 639
+[ �][�] -> [ आ] 640
+[ in][t] -> [ int] 641
+[ि�][�] -> [िए] 642
+[्�][�] -> [्त] 643
+[ou][t] -> [out] 644
+[ou][r] -> [our] 645
+[ u][s] -> [ us] 646
+[ p][l] -> [ pl] 647
+[u][st] -> [ust] 648
+[ I][n] -> [ In] 649
+[ the][ir] -> [ their] 650
+[ �][�] -> [ ಗ] 651
+[a][k] -> [ak] 652
+[ि�][�] -> [िल] 653
+[i][p] -> [ip] 654
+[ �][�] -> [ ಜ] 655
+[ಾಗ][ಿ] -> [ಾಗಿ] 656
+[ಂ][ತ] -> [ಂತ] 657
+[ the][y] -> [ they] 658
+[o][st] -> [ost] 659
+[ು][ತ್ತ] -> [ುತ್ತ] 660
+[ w][he] -> [ whe] 661
+[t][her] -> [ther] 662
+[ಾ�][�] -> [ಾಂ] 663
+[at][ed] -> [ated] 664
+[�][�] -> [ज] 665
+[ �][�] -> [ त] 666
+[ �][�] -> [ इ] 667
+[ ][im] -> [ im] 668
+[ 2][0] -> [ 20] 669
+[as][t] -> [ast] 670
+[್�][�] -> [್ಗ] 671
+[ �][�] -> [ श] 672
+[ि�][�] -> [िस] 673
+[ ][U] -> [ U] 674
+[ d][o] -> [ do] 675
+[ ][J] -> [ J] 676
+[ಿತ][ು] -> [ಿತು] 677
+[ ][en] -> [ en] 678
+[ಾ�][�] -> [ಾಲ] 679
+[ T][h] -> [ Th] 680
+[್�][�] -> [್ವ] 681
+[�][�] -> [ಜ] 682
+[ವ][ನ್ನು] -> [ವನ್ನು] 683
+[a][re] -> [are] 684
+[ou][s] -> [ous] 685
+[ थ][ा] -> [ था] 686
+[ you][r] -> [ your] 687
+[ al][l] -> [ all] 688
+[p][t] -> [pt] 689
+[a][ct] -> [act] 690
+[ion][s] -> [ions] 691
+[्�][�] -> [्व] 692
+[ m][e] -> [ me] 693
+[p][l] -> [pl] 694
+[f][f] -> [ff] 695
+[ wh][ich] -> [ which] 696
+[್�][�] -> [್ಸ] 697
+[ು][ದ] -> [ುದ] 698
+[on][g] -> [ong] 699
+[p][er] -> [per] 700
+[an][s] -> [ans] 701
+[e][ar] -> [ear] 702
+[ा�][�] -> [ाम] 703
+[ h][is] -> [ his] 704
+[ w][ill] -> [ will] 705
+[a][ke] -> [ake] 706
+[ b][ut] -> [ but] 707
+[ प][र] -> [ पर] 708
+[ �][�] -> [ ಒ] 709
+[ �][�] -> [ ದ] 710
+[ಿ�][�] -> [ಿಲ] 711
+[ g][o] -> [ go] 712
+[ h][as] -> [ has] 713
+[ag][e] -> [age] 714
+[id][e] -> [ide] 715
+[ m][ore] -> [ more] 716
+[c][l] -> [cl] 717
+[ा�][�] -> [ात] 718
+[ಿಗ][ೆ] -> [ಿಗೆ] 719
+[ಂ][ಡ] -> [ಂಡ] 720
+[त][ा] -> [ता] 721
+[ ल][िए] -> [ लिए] 722
+[d][u] -> [du] 723
+[ we][re] -> [ were] 724
+[�][�] -> [च] 725
+[in][d] -> [ind] 726
+[्�][�] -> [्ट] 727
+[k][s] -> [ks] 728
+[ m][an] -> [ man] 729
+[ com][p] -> [ comp] 730
+[ರ][್] -> [ರ್] 731
+[ि�][�] -> [िन] 732
+[i][b] -> [ib] 733
+[ re][s] -> [ res] 734
+[a][v] -> [av] 735
+[ क][िया] -> [ किया] 736
+[r][u] -> [ru] 737
+[ಿಸ][ಲ] -> [ಿಸಲ] 738
+[ಿದ][್ದ] -> [ಿದ್ದ] 739
+[ con][t] -> [ cont] 740
+[ab][le] -> [able] 741
+[ o][ut] -> [ out] 742
+[ent][s] -> [ents] 743
+[ ][j] -> [ j] 744
+[್�][�] -> [್ಮ] 745
+[ar][y] -> [ary] 746
+[r][y] -> [ry] 747
+[�][�] -> [ण] 748
+[i][z] -> [iz] 749
+[ವ][ು] -> [ವು] 750
+[ h][ad] -> [ had] 751
+[ ग�][�ा] -> [ गया] 752
+[ c][l] -> [ cl] 753
+[्�][�] -> [्ष] 754
+[ s][o] -> [ so] 755
+[ a][d] -> [ ad] 756
+[ he][r] -> [ her] 757
+[i][e] -> [ie] 758
+[ �][�] -> [ य] 759
+[or][m] -> [orm] 760
+[s][o] -> [so] 761
+[c][c] -> [cc] 762
+[ve][l] -> [vel] 763
+[ಿಂ][ದ] -> [ಿಂದ] 764
+[i][a] -> [ia] 765
+[as][s] -> [ass] 766
+[a][ch] -> [ach] 767
+[ न][े] -> [ ने] 768
+[ on][e] -> [ one] 769
+[ಂದ][ು] -> [ಂದು] 770
+[il][d] -> [ild] 771
+[ �][�] -> [ ख] 772
+[ I][t] -> [ It] 773
+[ ][3] -> [ 3] 774
+[ �][�] -> [ भ] 775
+[ �][�] -> [ च] 776
+[ul][t] -> [ult] 777
+[ �][�] -> [ ಉ] 778
+[i][re] -> [ire] 779
+[�][�] -> [ಧ] 780
+[ಾ�][�] -> [ಾಡ] 781
+[ಾ�][�] -> [ಾಸ] 782
+[o][o] -> [oo] 783
+[at][er] -> [ater] 784
+[ �][�] -> [ ಯ] 785
+[ o][ther] -> [ other] 786
+[ wh][o] -> [ who] 787
+[an][ce] -> [ance] 788
+[ �][�] -> [ फ] 789
+[ t][e] -> [ te] 790
+[on][e] -> [one] 791
+[en][ce] -> [ence] 792
+[े][ल] -> [ेल] 793
+[l][ow] -> [low] 794
+[ 2][00] -> [ 200] 795
+[ d][is] -> [ dis] 796
+[ation][s] -> [ations] 797
+[ಿದ][ೆ] -> [ಿದೆ] 798
+[a][ce] -> [ace] 799
+[�][�] -> [श] 800
+[b][er] -> [ber] 801
+[ा�][�] -> [ास] 802
+[am][e] -> [ame] 803
+[ದ][ಲ್ಲಿ] -> [ದಲ್ಲಿ] 804
+[ಾ�][�] -> [ಾವ] 805
+[p][ort] -> [port] 806
+[u][al] -> [ual] 807
+[ 1][8] -> [ 18] 808
+[ 20][1] -> [ 201] 809
+[re][at] -> [reat] 810
+[is][h] -> [ish] 811
+[n][ow] -> [now] 812
+[ the][m] -> [ them] 813
+[ t][im] -> [ tim] 814
+[್�][�] -> [್ಪ] 815
+[ al][so] -> [ also] 816
+[�][�] -> [ಭ] 817
+[ �][�] -> [ ಶ] 818
+[ s][a] -> [ sa] 819
+[an][g] -> [ang] 820
+[्�][�] -> [्स] 821
+[ver][y] -> [very] 822
+[ಕ್�][�] -> [ಕ್ಕ] 823
+[ उ][न] -> [ उन] 824
+[ u][n] -> [ un] 825
+[ ಸ][ಂ] -> [ ಸಂ] 826
+[್�][�] -> [್ಕ] 827
+[re][e] -> [ree] 828
+[ct][ion] -> [ction] 829
+[್�][�] -> [್ಥ] 830
+[ಗಳ][ು] -> [ಗಳು] 831
+[ic][e] -> [ice] 832
+[ �][�] -> [ ಚ] 833
+[ u][p] -> [ up] 834
+[ S][t] -> [ St] 835
+[ p][er] -> [ per] 836
+[ p][re] -> [ pre] 837
+[ಿಸ][ಿದ] -> [ಿಸಿದ] 838
+[ig][n] -> [ign] 839
+[�][�] -> [फ] 840
+[ स][ा�] -> [ सा�] 841
+[ा�][�] -> [ाद] 842
+[ೊ][ಂದ] -> [ೊಂದ] 843
+[ e][v] -> [ ev] 844
+[ ab][out] -> [ about] 845
+[�][�] -> [ಚ] 846
+[ the][re] -> [ there] 847
+[ಾ�][�] -> [ಾಮ] 848
+[ s][y] -> [ sy] 849
+[ p][art] -> [ part] 850
+[ac][k] -> [ack] 851
+[ a][r] -> [ ar] 852
+[ಗಳ][ನ್ನು] -> [ಗಳನ್ನು] 853
+[c][es] -> [ces] 854
+[ g][en] -> [ gen] 855
+[in][t] -> [int] 856
+[ರ][ೆ] -> [ರೆ] 857
+[�][�] -> [ध] 858
+[i][ke] -> [ike] 859
+[ व][ि�] -> [ वि�] 860
+[ou][nd] -> [ound] 861
+[ �][�] -> [ ड] 862
+[ Th][is] -> [ This] 863
+[ com][m] -> [ comm] 864
+[res][s] -> [ress] 865
+[it][e] -> [ite] 866
+[o][se] -> [ose] 867
+[ whe][n] -> [ when] 868
+[i][le] -> [ile] 869
+[a][u] -> [au] 870
+[ou][nt] -> [ount] 871
+[�][�] -> [ौ] 872
+[न][ा] -> [ना] 873
+[ int][o] -> [ into] 874
+[ s][c] -> [ sc] 875
+[ है][ं] -> [ हैं] 876
+[ �][�] -> [ ट] 877
+[ou][gh] -> [ough] 878
+[ स][ं�] -> [ सं�] 879
+[ s][pe] -> [ spe] 880
+[ s][he] -> [ she] 881
+[o][od] -> [ood] 882
+[ol][d] -> [old] 883
+[ू][प] -> [ूप] 884
+[ स][म] -> [ सम] 885
+[ �][�] -> [ ಲ] 886
+[ st][ud] -> [ stud] 887
+[ay][s] -> [ays] 888
+[ s][ome] -> [ some] 889
+[ th][an] -> [ than] 890
+[a][il] -> [ail] 891
+[as][e] -> [ase] 892
+[nd][er] -> [nder] 893
+[ p][e] -> [ pe] 894
+[ ಇ][ದ] -> [ ಇದ] 895
+[ಿ�][�] -> [ಿವ] 896
+[ಿಗ][ಳ] -> [ಿಗಳ] 897
+[if][f] -> [iff] 898
+[ ][Y] -> [ Y] 899
+[्�][�] -> [्ल] 900
+[u][e] -> [ue] 901
+[್�][�] -> [್ಚ] 902
+[ �][�] -> [ ಡ] 903
+[ಲ][್] -> [ಲ್] 904
+[ h][ow] -> [ how] 905
+[ s][p] -> [ sp] 906
+[e][w] -> [ew] 907
+[en][s] -> [ens] 908
+[u][b] -> [ub] 909
+[ n][o] -> [ no] 910
+[ be][en] -> [ been] 911
+[ t][w] -> [ tw] 912
+[ ][K] -> [ K] 913
+[ k][now] -> [ know] 914
+[it][ion] -> [ition] 915
+[ इ][स] -> [ इस] 916
+[ be][c] -> [ bec] 917
+[)][.] -> [).] 918
+[ಂ][ಬ] -> [ಂಬ] 919
+[ tim][e] -> [ time] 920
+[ y][ear] -> [ year] 921
+[ा�][�] -> [ां] 922
+[ ][4] -> [ 4] 923
+[k][ing] -> [king] 924
+[or][y] -> [ory] 925
+[e][p] -> [ep] 926
+[ ][if] -> [ if] 927
+[ o][ver] -> [ over] 928
+[ा�][�] -> [ाइ] 929
+[or][d] -> [ord] 930
+[ सा�][�] -> [ साथ] 931
+[ ][ra] -> [ ra] 932
+[ಗ][ೆ] -> [ಗೆ] 933
+[ ह][ो] -> [ हो] 934
+[ in][f] -> [ inf] 935
+[or][s] -> [ors] 936
+[ र][ूप] -> [ रूप] 937
+[ ][qu] -> [ qu] 938
+[ ಮ][ಾಡ] -> [ ಮಾಡ] 939
+[ f][ir] -> [ fir] 940
+[ಾಯ][ಿತು] -> [ಾಯಿತು] 941
+[ಾ�][�] -> [ಾಕ] 942
+[ा�][�] -> [ाप] 943
+[ d][iff] -> [ diff] 944
+[st][em] -> [stem] 945
+[i][an] -> [ian] 946
+[ C][h] -> [ Ch] 947
+[op][le] -> [ople] 948
+[ �][�] -> [ ಈ] 949
+[ a][g] -> [ ag] 950
+[त][्र] -> [त्र] 951
+[t][e] -> [te] 952
+[ l][ike] -> [ like] 953
+[w][n] -> [wn] 954
+[p][le] -> [ple] 955
+[ be][t] -> [ bet] 956
+[ w][ould] -> [ would] 957
+[ h][im] -> [ him] 958
+[ा�][�] -> [ाज] 959
+[on][d] -> [ond] 960
+[n][ing] -> [ning] 961
+[ a][pp] -> [ app] 962
+[ m][ay] -> [ may] 963
+[m][er] -> [mer] 964
+[ೇ][ರ] -> [ೇರ] 965
+[l][l] -> [ll] 966
+[ m][y] -> [ my] 967
+[rou][gh] -> [rough] 968
+[ ][ro] -> [ ro] 969
+[ ज][ो] -> [ जो] 970
+[�][�] -> [ೌ] 971
+[ic][k] -> [ick] 972
+[f][ter] -> [fter] 973
+[i][ous] -> [ious] 974
+[ an][y] -> [ any] 975
+[ದ][ೆ] -> [ದೆ] 976
+[ड][़] -> [ड़] 977
+[ wh][at] -> [ what] 978
+[c][re] -> [cre] 979
+[ಕ್�][�] -> [ಕ್ಷ] 980
+[ �][�] -> [ ಟ] 981
+[ अ][प] -> [ अप] 982
+[ವ][ಾಗಿ] -> [ವಾಗಿ] 983
+[ it][s] -> [ its] 984
+[er][v] -> [erv] 985
+[್ತ][ು] -> [್ತು] 986
+[ु][र] -> [ुर] 987
+[ಲ][ು] -> [ಲು] 988
+[o][ve] -> [ove] 989
+[ in][d] -> [ ind] 990
+[ो][ग] -> [ोग] 991
+[ o][ur] -> [ our] 992
+[र][ी] -> [री] 993
+[ u][se] -> [ use] 994
+[ of][f] -> [ off] 995
+[)][,] -> [),] 996
+[on][s] -> [ons] 997
+[ the][se] -> [ these] 998
+[ re][c] -> [ rec] 999
+[ ಅವ][ರು] -> [ ಅವರು] 1000
+[्�][�] -> [्ह] 1001
+[ing][s] -> [ings] 1002
+[ ][V] -> [ V] 1003
+[ pe][ople] -> [ people] 1004
+[re][n] -> [ren] 1005
+[ a][ct] -> [ act] 1006
+[�][�] -> [ಷ] 1007
+[ol][og] -> [olog] 1008
+[r][it] -> [rit] 1009
+[ u][nder] -> [ under] 1010
+[ಟ][್ಟ] -> [ಟ್ಟ] 1011
+[ A][s] -> [ As] 1012
+[ಾ�][�] -> [ಾಜ] 1013
+[p][h] -> [ph] 1014
+[w][e] -> [we] 1015
+[ d][es] -> [ des] 1016
+[ಿ�][�] -> [ಿಮ] 1017
+[ೇ][ಶ] -> [ೇಶ] 1018
+[ tw][o] -> [ two] 1019
+[ �][�] -> [ ಭ] 1020
+[ ][5] -> [ 5] 1021
+[ प्र][त] -> [ प्रत] 1022
+[al][ly] -> [ally] 1023
+[y][s] -> [ys] 1024
+[l][es] -> [les] 1025
+[se][l] -> [sel] 1026
+[ t][ra] -> [ tra] 1027
+[ೆ�][�] -> [ೆಸ] 1028
+[ he][l] -> [ hel] 1029
+[ा�][�] -> [ाय] 1030
+[is][s] -> [iss] 1031
+[j][ect] -> [ject] 1032
+[ sy][stem] -> [ system] 1033
+[ c][o] -> [ co] 1034
+[m][s] -> [ms] 1035
+[ा][र्�] -> [ार्�] 1036
+[ ne][w] -> [ new] 1037
+[ಂ][ಟ] -> [ಂಟ] 1038
+[le][ct] -> [lect] 1039
+[ic][t] -> [ict] 1040
+[ H][e] -> [ He] 1041
+[क][्ष] -> [क्ष] 1042
+[ re][l] -> [ rel] 1043
+[ hel][p] -> [ help] 1044
+[it][ies] -> [ities] 1045
+[au][se] -> [ause] 1046
+[्�][�] -> [्क] 1047
+[ar][k] -> [ark] 1048
+[at][es] -> [ates] 1049
+[ on][ly] -> [ only] 1050
+[ೊ][ಂಡ] -> [ೊಂಡ] 1051
+[ೆ][ಲ] -> [ೆಲ] 1052
+[ver][s] -> [vers] 1053
+[ m][ost] -> [ most] 1054
+[�][�] -> [भ] 1055
+[�][�] -> [ई] 1056
+[is][e] -> [ise] 1057
+[cl][ud] -> [clud] 1058
+[ ch][ild] -> [ child] 1059
+[ी][य] -> [ीय] 1060
+[ fir][st] -> [ first] 1061
+[ con][s] -> [ cons] 1062
+[ं�][�] -> [ंड] 1063
+[ t][r] -> [ tr] 1064
+[ W][h] -> [ Wh] 1065
+[ಂತ][ರ] -> [ಂತರ] 1066
+[ sh][ould] -> [ should] 1067
+[ul][ar] -> [ular] 1068
+[ ನ][ಿರ] -> [ ನಿರ] 1069
+[ಸ][್] -> [ಸ್] 1070
+[ ವ][ಿ�] -> [ ವಿ�] 1071
+[oo][k] -> [ook] 1072
+[f][ul] -> [ful] 1073
+[ುತ್ತ][ದೆ] -> [ುತ್ತದೆ] 1074
+[ವ][ಾಗ] -> [ವಾಗ] 1075
+[al][th] -> [alth] 1076
+[ man][y] -> [ many] 1077
+[ಕ್ಕ][ೆ] -> [ಕ್ಕೆ] 1078
+[ अ][न] -> [ अन] 1079
+[ा�][�] -> [ाव] 1080
+[್�][�] -> [್ಬ] 1081
+[ b][l] -> [ bl] 1082
+[�][�] -> [ಖ] 1083
+[|][|] -> [||] 1084
+[ p][o] -> [ po] 1085
+[ರ][ಣ] -> [ರಣ] 1086
+[�][�] -> [ೃ] 1087
+[ a][m] -> [ am] 1088
+[i][et] -> [iet] 1089
+[ಾ�][�] -> [ಾತ] 1090
+[ter][n] -> [tern] 1091
+[he][d] -> [hed] 1092
+[ प][ह] -> [ पह] 1093
+[ल][े] -> [ले] 1094
+[ c][ol] -> [ col] 1095
+[क][े] -> [के] 1096
+[ ][6] -> [ 6] 1097
+[�][�] -> [ए] 1098
+[ diff][ere] -> [ differe] 1099
+[ f][l] -> [ fl] 1100
+[1][9] -> [19] 1101
+[त][े] -> [ते] 1102
+[ m][od] -> [ mod] 1103
+[ ज][िस] -> [ जिस] 1104
+[ ಅವ][ರ] -> [ ಅವರ] 1105
+[ಾ�][�] -> [ಾಪ] 1106
+[c][ed] -> [ced] 1107
+[ spe][c] -> [ spec] 1108
+[ c][ould] -> [ could] 1109
+[if][e] -> [ife] 1110
+[್ಯ][ೂ] -> [್ಯೂ] 1111
+[c][ess] -> [cess] 1112
+[x][t] -> [xt] 1113
+[vel][op] -> [velop] 1114
+[ wor][k] -> [ work] 1115
+[�][�] -> [ृ] 1116
+[ कर][ने] -> [ करने] 1117
+[at][ing] -> [ating] 1118
+[ su][ch] -> [ such] 1119
+[ द][्व] -> [ द्व] 1120
+[if][ic] -> [ific] 1121
+[ೆಯ][ನ್ನು] -> [ೆಯನ್ನು] 1122
+[ಿತ][್ತು] -> [ಿತ್ತು] 1123
+[ f][e] -> [ fe] 1124
+[ार][ा] -> [ारा] 1125
+[ ][ ] -> [  ] 1126
+[ o][b] -> [ ob] 1127
+[f][ore] -> [fore] 1128
+[ �][�] -> [ ಫ] 1129
+[ उन][्ह] -> [ उन्ह] 1130
+[न][ी] -> [नी] 1131
+[re][d] -> [red] 1132
+[ff][ect] -> [ffect] 1133
+[at][ive] -> [ative] 1134
+[ೊ][ಳ] -> [ೊಳ] 1135
+[ थ][े] -> [ थे] 1136
+[ a][cc] -> [ acc] 1137
+[ p][r] -> [ pr] 1138
+[ th][rough] -> [ through] 1139
+[ w][ell] -> [ well] 1140
+[ಿರ][ುವ] -> [ಿರುವ] 1141
+[े�][�] -> [ेश] 1142
+[ s][ur] -> [ sur] 1143
+[i][ed] -> [ied] 1144
+[ g][et] -> [ get] 1145
+[ स्�][�] -> [ स्थ] 1146
+[he][n] -> [hen] 1147
+[o][v] -> [ov] 1148
+[er][t] -> [ert] 1149
+[ pro][du] -> [ produ] 1150
+[ a][fter] -> [ after] 1151
+[ sa][id] -> [ said] 1152
+[ pro][v] -> [ prov] 1153
+[ in][clud] -> [ includ] 1154
+[�][�] -> [ಫ] 1155
+[्�][�] -> [्प] 1156
+[ation][al] -> [ational] 1157
+[ure][s] -> [ures] 1158
+[os][s] -> [oss] 1159
+[स][े] -> [से] 1160
+[ p][h] -> [ ph] 1161
+[ थ][ी] -> [ थी] 1162
+[ w][ater] -> [ water] 1163
+[ the][n] -> [ then] 1164
+[ad][e] -> [ade] 1165
+[ನ][ು] -> [ನು] 1166
+[ re][m] -> [ rem] 1167
+[u][c] -> [uc] 1168
+[ow][n] -> [own] 1169
+[ur][n] -> [urn] 1170
+[್�][�] -> [್ಡ] 1171
+[ೇ][ಲ] -> [ೇಲ] 1172
+[t][ing] -> [ting] 1173
+[ot][h] -> [oth] 1174
+[b][le] -> [ble] 1175
+[ us][ed] -> [ used] 1176
+[at][h] -> [ath] 1177
+[್ಗ][ಳ] -> [್ಗಳ] 1178
+[ा�][�] -> [ाई] 1179
+[al][s] -> [als] 1180
+[and][s] -> [ands] 1181
+[ as][s] -> [ ass] 1182
+[ar][n] -> [arn] 1183
+[rou][nd] -> [round] 1184
+[ ಎ][ಂದು] -> [ ಎಂದು] 1185
+[l][ic] -> [lic] 1186
+[ R][e] -> [ Re] 1187
+[ ज][ात] -> [ जात] 1188
+[�][�] -> [उ] 1189
+[ re][g] -> [ reg] 1190
+[i][ent] -> [ient] 1191
+[in][k] -> [ink] 1192
+[ year][s] -> [ years] 1193
+[c][y] -> [cy] 1194
+[we][en] -> [ween] 1195
+[t][le] -> [tle] 1196
+[ de][velop] -> [ develop] 1197
+[ಿಸಿದ][ರು] -> [ಿಸಿದರು] 1198
+[ar][s] -> [ars] 1199
+[ ನ][ಂತರ] -> [ ನಂತರ] 1200
+[in][es] -> [ines] 1201
+[ಿಸ][ುವ] -> [ಿಸುವ] 1202
+[ U][n] -> [ Un] 1203
+[ ಕ][ಾರ] -> [ ಕಾರ] 1204
+[ೆ�][�] -> [ೆಚ] 1205
+[ि�][�] -> [िं] 1206
+[ द्व][ारा] -> [ द्वारा] 1207
+[ment][s] -> [ments] 1208
+[ಾ�][�] -> [ಾಣ] 1209
+[e][nd] -> [end] 1210
+[ri][b] -> [rib] 1211
+[ic][s] -> [ics] 1212
+[ s][m] -> [ sm] 1213
+[ su][b] -> [ sub] 1214
+[्�][�] -> [्म] 1215
+[ ಸ][ಮ] -> [ ಸಮ] 1216
+[ ne][ed] -> [ need] 1217
+[ भ][ी] -> [ भी] 1218
+[he][s] -> [hes] 1219
+[n][e] -> [ne] 1220
+[g][an] -> [gan] 1221
+[ su][pp] -> [ supp] 1222
+[u][ch] -> [uch] 1223
+[ur][ing] -> [uring] 1224
+[ವ][ೆ] -> [ವೆ] 1225
+[ र][ाज] -> [ राज] 1226
+[o][b] -> [ob] 1227
+[ ಇದ][ು] -> [ ಇದು] 1228
+[ೆಚ][್ಚ] -> [ೆಚ್ಚ] 1229
+[ ಸ][್ಥ] -> [ ಸ್ಥ] 1230
+[ whe][re] -> [ where] 1231
+[ im][p] -> [ imp] 1232
+[ar][ch] -> [arch] 1233
+[್�][�] -> [್ಧ] 1234
+[ು][ರ] -> [ುರ] 1235
+[ere][d] -> [ered] 1236
+[ a][c] -> [ ac] 1237
+[ व][र्�] -> [ वर्�] 1238
+[िय][ों] -> [ियों] 1239
+[ot][her] -> [other] 1240
+[ �][�] -> [ –] 1241
+[ bet][ween] -> [ between] 1242
+[an][ge] -> [ange] 1243
+[t][en] -> [ten] 1244
+[ at][t] -> [ att] 1245
+[್ಸ][್] -> [್ಸ್] 1246
+[ ಮ][ೂ] -> [ ಮೂ] 1247
+[ो][ल] -> [ोल] 1248
+[ स][ी] -> [ सी] 1249
+[ib][le] -> [ible] 1250
+[ೊಂದ][ಿಗೆ] -> [ೊಂದಿಗೆ] 1251
+[er][m] -> [erm] 1252
+[ಗಳ][ಲ್ಲಿ] -> [ಗಳಲ್ಲಿ] 1253
+[ in][v] -> [ inv] 1254
+[ 19][9] -> [ 199] 1255
+[ m][ake] -> [ make] 1256
+[2][0] -> [20] 1257
+[ अ][ध] -> [ अध] 1258
+[ य][ह] -> [ यह] 1259
+[ೆ][ರ] -> [ೆರ] 1260
+[ क][ि] -> [ कि] 1261
+[ ಮ][ು] -> [ ಮು] 1262
+[r][ic] -> [ric] 1263
+[ p][res] -> [ pres] 1264
+[ w][ay] -> [ way] 1265
+[et][h] -> [eth] 1266
+[ W][e] -> [ We] 1267
+[ ನ][ೀ] -> [ ನೀ] 1268
+[n][g] -> [ng] 1269
+[े][क] -> [ेक] 1270
+[ l][ong] -> [ long] 1271
+[ि�][�] -> [िम] 1272
+[े][न] -> [ेन] 1273
+[ in][ter] -> [ inter] 1274
+[ ad][d] -> [ add] 1275
+[ ][8] -> [ 8] 1276
+[ ज][ी] -> [ जी] 1277
+[ ex][am] -> [ exam] 1278
+[il][ity] -> [ility] 1279
+[p][s] -> [ps] 1280
+[ ह][ु�] -> [ हु�] 1281
+[it][s] -> [its] 1282
+[ಿಯ][ನ್] -> [ಿಯನ್] 1283
+[ro][w] -> [row] 1284
+[ l][ife] -> [ life] 1285
+[pe][ct] -> [pect] 1286
+[ for][m] -> [ form] 1287
+[ ][9] -> [ 9] 1288
+[ಿಲ][್ಲ] -> [ಿಲ್ಲ] 1289
+[ow][er] -> [ower] 1290
+[े�][�] -> [ेट] 1291
+[ f][ound] -> [ found] 1292
+[ उ][प] -> [ उप] 1293
+[as][h] -> [ash] 1294
+[ I][f] -> [ If] 1295
+[ु][ल] -> [ुल] 1296
+[�][�] -> [–] 1297
+[ gen][er] -> [ gener] 1298
+[ಂ][ಗ] -> [ಂಗ] 1299
+[ b][u] -> [ bu] 1300
+[ೆ][ಗೆ] -> [ೆಗೆ] 1301
+[o][y] -> [oy] 1302
+[as][ed] -> [ased] 1303
+[ in][st] -> [ inst] 1304
+[ j][ust] -> [ just] 1305
+[ ಪ][ರ] -> [ ಪರ] 1306
+[ ev][en] -> [ even] 1307
+[ differe][nt] -> [ different] 1308
+[ಿ�][�] -> [ಿಡ] 1309
+[ se][e] -> [ see] 1310
+[ ][7] -> [ 7] 1311
+[ द][िया] -> [ दिया] 1312
+[ ][ed] -> [ ed] 1313
+[ ख][ेल] -> [ खेल] 1314
+[ p][ers] -> [ pers] 1315
+[ H][ow] -> [ How] 1316
+[ c][all] -> [ call] 1317
+[in][s] -> [ins] 1318
+[�][�] -> [ಥ] 1319
+[ ಮ][ೊ] -> [ ಮೊ] 1320
+[e][ver] -> [ever] 1321
+[ों][ने] -> [ोंने] 1322
+[ o][wn] -> [ own] 1323
+[ ब][ाद] -> [ बाद] 1324
+[ th][ose] -> [ those] 1325
+[ e][m] -> [ em] 1326
+[ ಹ][ೆಚ್ಚ] -> [ ಹೆಚ್ಚ] 1327
+[ra][m] -> [ram] 1328
+[ch][n] -> [chn] 1329
+[ುದ][ು] -> [ುದು] 1330
+[ ಆ][ದ] -> [ ಆದ] 1331
+[ि�][�] -> [��व] 1332
+[sel][f] -> [self] 1333
+[mer][ic] -> [meric] 1334
+[ ][very] -> [ very] 1335
+[f][orm] -> [form] 1336
+[f][t] -> [ft] 1337
+[od][y] -> [ody] 1338
+[ n][um] -> [ num] 1339
+[ t][y] -> [ ty] 1340
+[उ][प] -> [उप] 1341
+[ೀ][ಯ] -> [ೀಯ] 1342
+[ tr][ans] -> [ trans] 1343
+[or][n] -> [orn] 1344
+[ಡ][್] -> [ಡ್] 1345
+[an][y] -> [any] 1346
+[ 1][0] -> [ 10] 1347
+[g][et] -> [get] 1348
+[ar][ly] -> [arly] 1349
+[o][ol] -> [ool] 1350
+[ ಹ][ೊಂದ] -> [ ಹೊಂದ] 1351
+[ant][s] -> [ants] 1352
+[್ನ][ಲ್ಲಿ] -> [್ನಲ್ಲಿ] 1353
+[ं][त] -> [ंत] 1354
+[ त][क] -> [ तक] 1355
+[a][w] -> [aw] 1356
+[ he][alth] -> [ health] 1357
+[r][on] -> [ron] 1358
+[ಿಸಲ][ು] -> [ಿಸಲು] 1359
+[ C][om] -> [ Com] 1360
+[ak][es] -> [akes] 1361
+[क][ार] -> [कार] 1362
+[ak][ing] -> [aking] 1363
+[ be][ing] -> [ being] 1364
+[ा�][�] -> [ाक] 1365
+[ in][cre] -> [ incre] 1366
+[र][ण] -> [रण] 1367
+[il][y] -> [ily] 1368
+[ e][very] -> [ every] 1369
+[ू][ल] -> [ूल] 1370
+[ ಹ][ೊ] -> [ ಹೊ] 1371
+[ಲ][್ಗ] -> [ಲ್ಗ] 1372
+[ ವ][್ಯ] -> [ ವ್ಯ] 1373
+[ಾ�][�] -> [ಾಟ] 1374
+[ to][o] -> [ too] 1375
+[t][y] -> [ty] 1376
+[�][्ट] -> [�्ट] 1377
+[ा�][�] -> [ाह] 1378
+[iv][es] -> [ives] 1379
+[1][0] -> [10] 1380
+[ra][ct] -> [ract] 1381
+[ical][ly] -> [ically] 1382
+[್ಯ][ಾಂ] -> [್ಯಾಂ] 1383
+[ d][et] -> [ det] 1384
+[ P][ro] -> [ Pro] 1385
+[ h][um] -> [ hum] 1386
+[t][s] -> [ts] 1387
+[ h][igh] -> [ high] 1388
+[ ಸ][ಹ] -> [ ಸಹ] 1389
+[ ಅ][ನ್ನು] -> [ ಅನ್ನು] 1390
+[c][k] -> [ck] 1391
+[ ][ent] -> [ ent] 1392
+[ The][re] -> [ There] 1393
+[ಾ�][�] -> [ಾಹ] 1394
+[ಸ][್ತ] -> [ಸ್ತ] 1395
+[ A][meric] -> [ Americ] 1396
+[ l][ar] -> [ lar] 1397
+[ be][fore] -> [ before] 1398
+[ವ][ಾದ] -> [ವಾದ] 1399
+[00][0] -> [000] 1400
+[ह][ी] -> [ही] 1401
+[ ex][p] -> [ exp] 1402
+[ B][ut] -> [ But] 1403
+[क][्त] -> [क्त] 1404
+[ e][ach] -> [ each] 1405
+[ ex][per] -> [ exper] 1406
+[oc][k] -> [ock] 1407
+[ ಅ][ದ] -> [ ಅದ] 1408
+[i][ence] -> [ience] 1409
+[ m][uch] -> [ much] 1410
+[ist][s] -> [ists] 1411
+[ व][ाल] -> [ वाल] 1412
+[id][ent] -> [ident] 1413
+[ಿ�][�] -> [ಿಟ] 1414
+[n][ess] -> [ness] 1415
+[ c][ount] -> [ count] 1416
+[ s][im] -> [ sim] 1417
+[ ][ke] -> [ ke] 1418
+[ S][e] -> [ Se] 1419
+[ar][m] -> [arm] 1420
+[ ಜ][ನ] -> [ ಜನ] 1421
+[ा�][�] -> [ाग] 1422
+[ut][ion] -> [ution] 1423
+[ಗ][್] -> [ಗ್] 1424
+[ि][र्�] -> [िर्�] 1425
+[ f][ol] -> [ fol] 1426
+[ e][nd] -> [ end] 1427
+[er][g] -> [erg] 1428
+[ಡ][ೆ] -> [ಡೆ] 1429
+[ ][|] -> [ |] 1430
+[ im][port] -> [ import] 1431
+[ de][c] -> [ dec] 1432
+[ child][ren] -> [ children] 1433
+[ f][ind] -> [ find] 1434
+[ re][f] -> [ ref] 1435
+[ wor][ld] -> [ world] 1436
+[ The][y] -> [ They] 1437
+[ನ][ೆಯ] -> [ನೆಯ] 1438
+[ೆ�][�] -> [ೆಟ] 1439
+[ज][़] -> [ज़] 1440
+[ re][ad] -> [ read] 1441
+[ m][ade] -> [ made] 1442
+[ದ][ಲ] -> [ದಲ] 1443
+[ राज][्य] -> [ राज्य] 1444
+[ do][es] -> [ does] 1445
+[ual][ly] -> [ually] 1446
+[ s][oc] -> [ soc] 1447
+[olog][y] -> [ology] 1448
+[ A][l] -> [ Al] 1449
+[ n][ow] -> [ now] 1450
+[ಿದ][ರು] -> [ಿದರು] 1451
+[े][र] -> [ेर] 1452
+[ं�][�] -> [ंद] 1453
+[ go][od] -> [ good] 1454
+[ जात][ा] -> [ जाता] 1455
+[ಾನ][್] -> [ಾನ್] 1456
+[ s][k] -> [ sk] 1457
+[ d][own] -> [ down] 1458
+[er][n] -> [ern] 1459
+[ e][ffect] -> [ effect] 1460
+[at][a] -> [ata] 1461
+[ೇ][ಕ] -> [ೇಕ] 1462
+[ उन्ह][ोंने] -> [ उन्होंने] 1463
+[त][ी] -> [ती] 1464
+[ c][re] -> [ cre] 1465
+[en][e] -> [ene] 1466
+[ೋ][ಜ] -> [ೋಜ] 1467
+[ू][र्�] -> [ूर्�] 1468
+[ಿ�][�] -> [ಿಷ] 1469
+[at][ure] -> [ature] 1470
+[l][ed] -> [led] 1471
+[ರ][ಡ] -> [ರಡ] 1472
+[v][ent] -> [vent] 1473
+[ur][al] -> [ural] 1474
+[ಾರ][ೆ] -> [ಾರೆ] 1475
+[ re][t] -> [ ret] 1476
+[म][ें] -> [में] 1477
+[ಕ][್ರ] -> [ಕ್ರ] 1478
+[ ವ][ರ್�] -> [ ವರ್�] 1479
+[ल][ी] -> [ली] 1480
+[ि�][�] -> [िश] 1481
+[ t][ake] -> [ take] 1482
+[ st][r] -> [ str] 1483
+[ b][ack] -> [ back] 1484
+[ F][or] -> [ For] 1485
+[ಾಗ][ಿದ್ದ] -> [ಾಗಿದ್ದ] 1486
+[ C][l] -> [ Cl] 1487
+[ಿಯ][ಲ್ಲಿ] -> [ಿಯಲ್ಲಿ] 1488
+[ t][ri] -> [ tri] 1489
+[ಿ�][�] -> [ಿಪ] 1490
+[ಸ][್ಥ] -> [ಸ್ಥ] 1491
+[ स्�][�] -> [ स्ट] 1492
+[ अध][िक] -> [ अधिक] 1493
+[ re][se] -> [ rese] 1494
+[ l][it] -> [ lit] 1495
+[ag][es] -> [ages] 1496
+[it][ed] -> [ited] 1497
+[ bec][ause] -> [ because] 1498
+[ ब][न] -> [ बन] 1499
+[ द][ो] -> [ दो] 1500
+[ c][ent] -> [ cent] 1501
+[ th][ough] -> [ though] 1502
+[ou][n] -> [oun] 1503
+[ ರ][ಾಜ] -> [ ರಾಜ] 1504
+[ ಮ][ೇಲ] -> [ ಮೇಲ] 1505
+[ v][ar] -> [ var] 1506
+[ fol][low] -> [ follow] 1507
+[ स्�][�] -> [ स्व] 1508
+[ c][le] -> [ cle] 1509
+[ m][on] -> [ mon] 1510
+[ ag][ain] -> [ again] 1511
+[ ಪ್ರ][ತ] -> [ ಪ್ರತ] 1512
+[ಿಸಲ][ಾಯಿತು] -> [ಿಸಲಾಯಿತು] 1513
+[े][स] -> [ेस] 1514
+[ल][ा] -> [ला] 1515
+[v][ed] -> [ved] 1516
+[v][es] -> [ves] 1517
+[ d][ay] -> [ day] 1518
+[ce][pt] -> [cept] 1519
+[ l][and] -> [ land] 1520
+[at][her] -> [ather] 1521
+[ ಮ][ಾರ] -> [ ಮಾರ] 1522
+[l][and] -> [land] 1523
+[ನ][ೆ] -> [ನೆ] 1524
+[ell][s] -> [ells] 1525
+[ inf][orm] -> [ inform] 1526
+[ m][ed] -> [ med] 1527
+[ु�][�] -> [ुद] 1528
+[ಂ][ಭ] -> [ಂಭ] 1529
+[ु][न] -> [ुन] 1530
+[in][ce] -> [ince] 1531
+[c][on] -> [con] 1532
+[ru][ct] -> [ruct] 1533
+[ಿಯ][ನ್ನು] -> [ಿಯನ್ನು] 1534
+[ि][र] -> [िर] 1535
+[ ಸ][ಾಮ] -> [ ಸಾಮ] 1536
+[ch][ool] -> [chool] 1537
+[ं�][�] -> [ंग] 1538
+[र्�][�] -> [र्य] 1539
+[w][ard] -> [ward] 1540
+[ te][chn] -> [ techn] 1541
+[ic][es] -> [ices] 1542
+[ न][िर्�] -> [ निर्�] 1543
+[ f][act] -> [ fact] 1544
+[ s][ame] -> [ same] 1545
+[ d][id] -> [ did] 1546
+[ೂ][ರ್�] -> [ೂರ್�] 1547
+[ cl][ass] -> [ class] 1548
+[क][ी] -> [की] 1549
+[ A][n] -> [ An] 1550
+[ೋ][ಗ] -> [ೋಗ] 1551
+[ಮ][್ಮ] -> [ಮ್ಮ] 1552
+[र्�][�] -> [र्ड] 1553
+[a][ir] -> [air] 1554
+[c][ial] -> [cial] 1555
+[ub][lic] -> [ublic] 1556
+[ra][ph] -> [raph] 1557
+[क][ि] -> [कि] 1558
+[ ಕ][್ರ] -> [ ಕ್ರ] 1559
+[a][j] -> [aj] 1560
+[ res][ult] -> [ result] 1561
+[िं][ग] -> [िंग] 1562
+[ಾಗ][ಿದೆ] -> [ಾಗಿದೆ] 1563
+[ pro][cess] -> [ process] 1564
+[ pers][on] -> [ person] 1565
+[i][vers] -> [ivers] 1566
+[ s][et] -> [ set] 1567
+[ b][oth] -> [ both] 1568
+[ in][s] -> [ ins] 1569
+[ು][ಗ] -> [ುಗ] 1570
+[ ex][pl] -> [ expl] 1571
+[a][h] -> [ah] 1572
+[iz][ed] -> [ized] 1573
+[ b][ody] -> [ body] 1574
+[स][्क] -> [स्क] 1575
+[ अप][ने] -> [ अपने] 1576
+[T][he] -> [The] 1577
+[d][s] -> [ds] 1578
+[ಕ][ರ] -> [ಕರ] 1579
+[ॉ][ल] -> [ॉल] 1580
+[d][er] -> [der] 1581
+[स][्त] -> [स्त] 1582
+[ W][hen] -> [ When] 1583
+[ s][erv] -> [ serv] 1584
+[w][ay] -> [way] 1585
+[ी][न] -> [ीन] 1586
+[ವ][ಾ] -> [ವಾ] 1587
+[ ಒ][ಂದು] -> [ ಒಂದು] 1588
+[ stud][ents] -> [ students] 1589
+[ ಬ][ಲ್ಗ] -> [ ಬಲ್ಗ] 1590
+[u][es] -> [ues] 1591
+[ th][ree] -> [ three] 1592
+[ �][�] -> [ ಏ] 1593
+[ ಆ][ಗ] -> [ ಆಗ] 1594
+[ro][s] -> [ros] 1595
+[ ब][ी] -> [ बी] 1596
+[ार][ी] -> [ारी] 1597
+[ stud][y] -> [ study] 1598
+[ dis][c] -> [ disc] 1599
+[ sm][all] -> [ small] 1600
+[ m][ight] -> [ might] 1601
+[ o][p] -> [ op] 1602
+[ re][p] -> [ rep] 1603
+[i][x] -> [ix] 1604
+[्य][ू] -> [्यू] 1605
+[ d][ist] -> [ dist] 1606
+[ c][ar] -> [ car] 1607
+[v][ing] -> [ving] 1608
+[್ಟ][್ರ] -> [್ಟ್ರ] 1609
+[ pro][g] -> [ prog] 1610
+[ा�][�] -> [ाँ] 1611
+[ p][at] -> [ pat] 1612
+[ import][ant] -> [ important] 1613
+[t][ed] -> [ted] 1614
+[ comm][un] -> [ commun] 1615
+[ ಎ][ರಡ] -> [ ಎರಡ] 1616
+[ಿಯ][ಾದ] -> [ಿಯಾದ] 1617
+[ वि�][�] -> [ विश] 1618
+[ �][�] -> [ घ] 1619
+[ S][he] -> [ She] 1620
+[ अ][ंत] -> [ अंत] 1621
+[ wh][ile] -> [ while] 1622
+[ O][n] -> [ On] 1623
+[ b][o] -> [ bo] 1624
+[ de][f] -> [ def] 1625
+[ g][rou] -> [ grou] 1626
+[�][ं] -> [�ं] 1627
+[ c][ap] -> [ cap] 1628
+[ re][qu] -> [ requ] 1629
+[ क][्ष] -> [ क्ष] 1630
+[ा�][�] -> [ाड] 1631
+[ c][ur] -> [ cur] 1632
+[ ಅ][ಧ] -> [ ಅಧ] 1633
+[ं][प] -> [ंप] 1634
+[ h][and] -> [ hand] 1635
+[ow][s] -> [ows] 1636
+[ E][th] -> [ Eth] 1637
+[st][and] -> [stand] 1638
+[ c][he] -> [ che] 1639
+[ 19][7] -> [ 197] 1640
+[ು][ಕ] -> [ುಕ] 1641
+[ en][g] -> [ eng] 1642
+[ro][ad] -> [road] 1643
+[ are][a] -> [ area] 1644
+[ವ][ರ] -> [ವರ] 1645
+[g][en] -> [gen] 1646
+[ f][low] -> [ flow] 1647
+[ st][art] -> [ start] 1648
+[ म][ै] -> [ मै] 1649
+[ान][े] -> [ाने] 1650
+[्�][�] -> [्न] 1651
+[ le][vel] -> [ level] 1652
+[ f][am] -> [ fam] 1653
+[ ಮೂ][ಲ] -> [ ಮೂಲ] 1654
+[ि�][�] -> [िट] 1655
+[ l][ook] -> [ look] 1656
+[ po][int] -> [ point] 1657
+[್�][�] -> [್ಳ] 1658
+[े�][�] -> [ेव] 1659
+[�][�] -> [ख] 1660
+[id][es] -> [ides] 1661
+[ ar][t] -> [ art] 1662
+[ comp][le] -> [ comple] 1663
+[i][o] -> [io] 1664
+[ ಉ][ದ] -> [ ಉದ] 1665
+[ m][ust] -> [ must] 1666
+[ us][ing] -> [ using] 1667
+[ d][uring] -> [ during] 1668
+[ie][w] -> [iew] 1669
+[d][ay] -> [day] 1670
+[ ನ][ಡ] -> [ ನಡ] 1671
+[ ch][ar] -> [ char] 1672
+[ part][ic] -> [ partic] 1673
+[ �][�] -> [ छ] 1674
+[ p][oss] -> [ poss] 1675
+[ p][ol] -> [ pol] 1676
+[ ಪ್ರ][ದ] -> [ ಪ್ರದ] 1677
+[್ಗ][ೆ] -> [್ಗೆ] 1678
+[ g][ra] -> [ gra] 1679
+[ s][ign] -> [ sign] 1680
+[ b][r] -> [ br] 1681
+[ 1][7] -> [ 17] 1682
+[v][en] -> [ven] 1683
+[ pl][ace] -> [ place] 1684
+[ v][al] -> [ val] 1685
+[ತ][ರ] -> [ತರ] 1686
+[ೆ][ನ] -> [ೆನ] 1687
+[ि�][�] -> [िप] 1688
+[ही][ं] -> [हीं] 1689
+[ಿಕ][್] -> [ಿಕ್] 1690
+[े][म] -> [ेम] 1691
+[ exam][ple] -> [ example] 1692
+[क][ा] -> [का] 1693
+[ उ][त] -> [ उत] 1694
+[ g][reat] -> [ great] 1695
+[य][ोग] -> [योग] 1696
+[c][ent] -> [cent] 1697
+[ ग�][�] -> [ गए] 1698
+[h][ip] -> [hip] 1699
+[ m][in] -> [ min] 1700
+[ ಅ][ಥ] -> [ ಅಥ] 1701
+[ಗ][ೊಂಡ] -> [ಗೊಂಡ] 1702
+[g][er] -> [ger] 1703
+[ि�][�] -> [िज] 1704
+[re][ad] -> [read] 1705
+[ain][s] -> [ains] 1706
+[iet][y] -> [iety] 1707
+[ ನೀ][ಡ] -> [ ನೀಡ] 1708
+[ौ][र] -> [ौर] 1709
+[ter][s] -> [ters] 1710
+[ or][gan] -> [ organ] 1711
+[ pl][an] -> [ plan] 1712
+[ f][un] -> [ fun] 1713
+[ व][्य] -> [ व्य] 1714
+[ाप][्त] -> [ाप्त] 1715
+[ು][ನ] -> [ುನ] 1716
+[ B][e] -> [ Be] 1717
+[ತ][್ಯ] -> [ತ್ಯ] 1718
+[ r][ight] -> [ right] 1719
+[iz][e] -> [ize] 1720
+[ an][other] -> [ another] 1721
+[ 19][8] -> [ 198] 1722
+[ e][lect] -> [ elect] 1723
+[ a][p] -> [ ap] 1724
+[ d][ata] -> [ data] 1725
+[ ][est] -> [ est] 1726
+[ं�][�] -> [ंट] 1727
+[ न][हीं] -> [ नहीं] 1728
+[ f][ree] -> [ free] 1729
+[क][्र] -> [क्र] 1730
+[ ಸ][ೇ] -> [ ಸೇ] 1731
+[g][g] -> [gg] 1732
+[ಿಸ][ುತ್ತ] -> [ಿಸುತ್ತ] 1733
+[ou][se] -> [ouse] 1734
+[in][ed] -> [ined] 1735
+[n][ot] -> [not] 1736
+[ut][e] -> [ute] 1737
+[्�][�] -> [्ध] 1738
+[re][en] -> [reen] 1739
+[ Y][ou] -> [ You] 1740
+[ app][ro] -> [ appro] 1741
+[ರ್�][�] -> [ರ್ಗ] 1742
+[ ಗ][ು] -> [ ಗು] 1743
+[ des][ign] -> [ design] 1744
+[ A][r] -> [ Ar] 1745
+[is][m] -> [ism] 1746
+[ೋ][ರ] -> [ೋರ] 1747
+[ s][om] -> [ som] 1748
+[ st][ill] -> [ still] 1749
+[ S][c] -> [ Sc] 1750
+[ो][र] -> [ोर] 1751
+[ con][f] -> [ conf] 1752
+[ D][e] -> [ De] 1753
+[े�][�] -> [ेज] 1754
+[ a][ut] -> [ aut] 1755
+[ar][th] -> [arth] 1756
+[ೆಯ][ು] -> [ೆಯು] 1757
+[ र][ह] -> [ रह] 1758
+[id][s] -> [ids] 1759
+[a][ve] -> [ave] 1760
+[ी][त] -> [ीत] 1761
+[iv][id] -> [ivid] 1762
+[ो][ट] -> [ोट] 1763
+[ान][ा] -> [ाना] 1764
+[ ತ][ನ್ನ] -> [ ತನ್ನ] 1765
+[ಾ�][�] -> [ಾಷ] 1766
+[ ಪ][ೂರ್�] -> [ ಪೂರ್�] 1767
+[ ಕ][ೊ] -> [ ಕೊ] 1768
+[ ಬಲ್ಗ][ೇರ] -> [ ಬಲ್ಗೇರ] 1769
+[ ex][t] -> [ ext] 1770
+[ of][ten] -> [ often] 1771
+[ा�][�ा] -> [ाया] 1772
+[ ಗ][್ರ] -> [ ಗ್ರ] 1773
+[ m][ain] -> [ main] 1774
+[ 1][2] -> [ 12] 1775
+[ य][ू] -> [ यू] 1776
+[ा�][�ं] -> [ाओं] 1777
+[ con][st] -> [ const] 1778
+[ l][oc] -> [ loc] 1779
+[ ಅಥ][ವಾ] -> [ ಅಥವಾ] 1780
+[ಿಸ][ಿತು] -> [ಿಸಿತು] 1781
+[ಪ][್ರ] -> [ಪ್ರ] 1782
+[ ][.] -> [ .] 1783
+[are][d] -> [ared] 1784
+[erg][y] -> [ergy] 1785
+[ का][र्य] -> [ कार्य] 1786
+[ 1][5] -> [ 15] 1787
+[er][ing] -> [ering] 1788
+[ pro][ble] -> [ proble] 1789
+[ ch][ang] -> [ chang] 1790
+[ s][ol] -> [ sol] 1791
+[ Wh][at] -> [ What] 1792
+[ num][ber] -> [ number] 1793
+[ಪ][್ಪ] -> [ಪ್ಪ] 1794
+[ a][v] -> [ av] 1795
+[ inform][ation] -> [ information] 1796
+[ w][om] -> [ wom] 1797
+[i][er] -> [ier] 1798
+[ क][ार] -> [ कार] 1799
+[act][er] -> [acter] 1800
+[ cons][id] -> [ consid] 1801
+[ T][o] -> [ To] 1802
+[ s][chool] -> [ school] 1803
+[it][al] -> [ital] 1804
+[ ಕಾರ][್ಯ] -> [ ಕಾರ್ಯ] 1805
+[ ]['] -> [ '] 1806
+[o][int] -> [oint] 1807
+[ರ್�][�] -> [ರ್ಶ] 1808
+[ ed][uc] -> [ educ] 1809
+[og][raph] -> [ograph] 1810
+[ d][ire] -> [ dire] 1811
+[े][त्र] -> [ेत्र] 1812
+[ ಭ][ಾಗ] -> [ ಭಾಗ] 1813
+[ಕ][ಾರ] -> [ಕಾರ] 1814
+[ C][on] -> [ Con] 1815
+[ A][nd] -> [ And] 1816
+[c][om] -> [com] 1817
+[ 19][4] -> [ 194] 1818
+[ le][ad] -> [ lead] 1819
+[�][�] -> [ಎ] 1820
+[ಾ�][�] -> [ಾಧ] 1821
+[ स्�][�] -> [ स्क] 1822
+[in][al] -> [inal] 1823
+[ l][ess] -> [ less] 1824
+[ pro][t] -> [ prot] 1825
+[ o][ld] -> [ old] 1826
+[em][ent] -> [ement] 1827
+[ v][is] -> [ vis] 1828
+[ f][ew] -> [ few] 1829
+[ 19][6] -> [ 196] 1830
+[ಖ][್ಯ] -> [ಖ್ಯ] 1831
+[ृ][त] -> [ृत] 1832
+[ a][round] -> [ around] 1833
+[ ][ve] -> [ ve] 1834
+[ s][er] -> [ ser] 1835
+[ಿಗಳ][ು] -> [ಿಗಳು] 1836
+[ಂ][ಪ] -> [ಂಪ] 1837
+[ पह][ले] -> [ पहले] 1838
+[ff][ic] -> [ffic] 1839
+[ g][l] -> [ gl] 1840
+[ದ][ಿಂದ] -> [ದಿಂದ] 1841
+[ೋ][ಪ] -> [ೋಪ] 1842
+[ वर्�][�] -> [ वर्ष] 1843
+[ pl][ay] -> [ play] 1844
+[ ಬ][ಳ] -> [ ಬಳ] 1845
+[ಾಂ][ಟ] -> [ಾಂಟ] 1846
+[े�][�] -> [ेड] 1847
+[ate][ly] -> [ately] 1848
+[ pro][f] -> [ prof] 1849
+[u][ro] -> [uro] 1850
+[ p][ower] -> [ power] 1851
+[ श][ुर] -> [ ���ुर] 1852
+[ स्थ][ान] -> [ स्थान] 1853
+[ೀ][ವ] -> [ೀವ] 1854
+[el][s] -> [els] 1855
+[ P][h] -> [ Ph] 1856
+[ क्ष][ेत्र] -> [ क्षेत्र] 1857
+[ स][ं] -> [ सं] 1858
+[स][ी] -> [सी] 1859
+[ o][cc] -> [ occ] 1860
+[ Eth][an] -> [ Ethan] 1861
+[ h][ist] -> [ hist] 1862
+[ಟ][್] -> [ಟ್] 1863
+[ प][ुर] -> [ पुर] 1864
+[o][ot] -> [oot] 1865
+[i][en] -> [ien] 1866
+[her][s] -> [hers] 1867
+[u][nd] -> [und] 1868
+[ re][port] -> [ report] 1869
+[ �][�] -> [ ಖ] 1870
+[ೆಯ][ಲ್ಲಿ] -> [ೆಯಲ್ಲಿ] 1871
+[ act][iv] -> [ activ] 1872
+[ w][rit] -> [ writ] 1873
+[ श][ाम] -> [ शाम] 1874
+[ul][l] -> [ull] 1875
+[ c][a] -> [ ca] 1876
+[he][re] -> [here] 1877
+[ d][i] -> [ di] 1878
+[ೊಳ][್ಳ] -> [ೊಳ್ಳ] 1879
+[ver][n] -> [vern] 1880
+[r][al] -> [ral] 1881
+[as][es] -> [ases] 1882
+[1][2] -> [12] 1883
+[ere][st] -> [erest] 1884
+[ರ್�][�] -> [ರ್ಕ] 1885
+[ ಕ][ೆಲ] -> [ ಕೆಲ] 1886
+[ ज][ब] -> [ जब] 1887
+[स][्ट] -> [स्ट] 1888
+[ l][ast] -> [ last] 1889
+[ m][us] -> [ mus] 1890
+[or][k] -> [ork] 1891
+[ं�][�] -> [ंब] 1892
+[ se][c] -> [ sec] 1893
+[ comm][on] -> [ common] 1894
+[pe][nd] -> [pend] 1895
+[ be][l] -> [ bel] 1896
+[ ಮೊ][ದಲ] -> [ ಮೊದಲ] 1897
+[ call][ed] -> [ called] 1898
+[ೇ][ಳ] -> [ೇಳ] 1899
+[ pres][ent] -> [ present] 1900
+[ st][and] -> [ stand] 1901
+[b][o] -> [bo] 1902
+[ou][th] -> [outh] 1903
+[ p][ass] -> [ pass] 1904
+[ ][<] -> [ <] 1905
+[ v][ir] -> [ vir] 1906
+[ A][t] -> [ At] 1907
+[ಾರ][ು] -> [ಾರು] 1908
+[ಿ�][�] -> [ಿಶ] 1909
+[ ಎ][ಂಬ] -> [ ಎಂಬ] 1910
+[ सं�][�] -> [ संग] 1911
+[ r][is] -> [ ris] 1912
+[ know][n] -> [ known] 1913
+[ U][S] -> [ US] 1914
+[al][e] -> [ale] 1915
+[ri][es] -> [ries] 1916
+[ c][or] -> [ cor] 1917
+[ The][se] -> [ These] 1918
+[ ಸ][್ವ] -> [ ಸ್ವ] 1919
+[ under][stand] -> [ understand] 1920
+[्�][�] -> [्च] 1921
+[ bu][ild] -> [ build] 1922
+[ ][[] -> [ [] 1923
+[ शाम][िल] -> [ शामिल] 1924
+[ert][ain] -> [ertain] 1925
+[ op][en] -> [ open] 1926
+[ ಮೇಲ][ೆ] -> [ ಮೇಲೆ] 1927
+[ಷ][್ಟ] -> [ಷ್ಟ] 1928
+[ ट][ी] -> [ टी] 1929
+[ अन][ु] -> [ अनु] 1930
+[en][a] -> [ena] 1931
+[ be][g] -> [ beg] 1932
+[ಿ�][�] -> [ಿಣ] 1933
+[िय][ो] -> [ियो] 1934
+[ E][x] -> [ Ex] 1935
+[ res][p] -> [ resp] 1936
+[ ಅ][ನ] -> [ ಅನ] 1937
+[ s][ay] -> [ say] 1938
+[ा�][�] -> [ाध] 1939
+[ e][qu] -> [ equ] 1940
+[ೆ][ಂಟ] -> [ೆಂಟ] 1941
+[w][ays] -> [ways] 1942
+[ supp][ort] -> [ support] 1943
+[et][s] -> [ets] 1944
+[ c][our] -> [ cour] 1945
+[ d][on] -> [ don] 1946
+[ु�][�] -> [ुख] 1947
+[is][hed] -> [ished] 1948
+[ lar][ge] -> [ large] 1949
+[ E][ng] -> [ Eng] 1950
+[ 1][6] -> [ 16] 1951
+[ ][0] -> [ 0] 1952
+[ �][�] -> [ ऑ] 1953
+[ w][ant] -> [ want] 1954
+[ with][out] -> [ without] 1955
+[ p][ain] -> [ pain] 1956
+[ h][y] -> [ hy] 1957
+[ h][ome] -> [ home] 1958
+[ e][as] -> [ eas] 1959
+[ire][d] -> [ired] 1960
+[್�][�] -> [್ಷ] 1961
+[ ಒ][ಳ] -> [ ಒಳ] 1962
+[ क][ै] -> [ कै] 1963
+[ ನಿರ][್ಮ] -> [ ನಿರ್ಮ] 1964
+[ ವರ್�][�] -> [ ವರ್ಷ] 1965
+[ರ][ದ] -> [ರದ] 1966
+[ 1][4] -> [ 14] 1967
+[ा�][�्ट] -> [ाष्ट] 1968
+[ाष्ट][्र] -> [ाष्ट्र] 1969
+[ा�][�] -> [ाए] 1970
+[ ग�][�] -> [ गई] 1971
+[ं][त्र] -> [ंत्र] 1972
+[ t][em] -> [ tem] 1973
+[v][ir] -> [vir] 1974
+[ t][reat] -> [ treat] 1975
+[ನ][ೇ] -> [ನೇ] 1976
+[ೋ][ಲ] -> [ೋಲ] 1977
+[ಿಕ][ೊಂಡ] -> [ಿಕೊಂಡ] 1978
+[ ख][िल] -> [ खिल] 1979
+[ lit][tle] -> [ little] 1980
+[ bet][ter] -> [ better] 1981
+[ c][are] -> [ care] 1982
+[ p][op] -> [ pop] 1983
+[in][ing] -> [ining] 1984
+[i][el] -> [iel] 1985
+[ ಆ][ರ] -> [ ಆರ] 1986
+[ि�][�] -> [िग] 1987
+[न][ों] -> [नों] 1988
+[u][p] -> [up] 1989
+[ros][s] -> [ross] 1990
+[ಲ್ಲ][ಿರುವ] -> [ಲ್ಲಿರುವ] 1991
+[un][k] -> [unk] 1992
+[h][old] -> [hold] 1993
+[ivers][ity] -> [iversity] 1994
+[ im][m] -> [ imm] 1995
+[्�][�] -> [्थ] 1996
+[p][r] -> [pr] 1997
+[ hum][an] -> [ human] 1998
+[iss][ion] -> [ission] 1999
+[ill][s] -> [ills] 2000
+[ with][in] -> [ within] 2001
+[ ತ][ಮ್ಮ] -> [ ತಮ್ಮ] 2002
+[भ][ी] -> [भी] 2003
+[ �][�ा] -> [ या] 2004
+[d][ition] -> [dition] 2005
+[ाल][य] -> [ालय] 2006
+[ l][ight] -> [ light] 2007
+[ ए][फ] -> [ एफ] 2008
+[ा�][�] -> [ाउ] 2009
+[ न][ाम] -> [ नाम] 2010
+[iv][ing] -> [iving] 2011
+[ int][erest] -> [ interest] 2012
+[ಲ][ಾಯಿತು] -> [ಲಾಯಿತು] 2013
+[ p][ublic] -> [ public] 2014
+[ I][nd] -> [ Ind] 2015
+[p][es] -> [pes] 2016
+[ C][ol] -> [ Col] 2017
+[ per][form] -> [ perform] 2018
+[ಕ್�][�] -> [ಕ್ಸ] 2019
+[ಾತ][್ರ] -> [ಾತ್ರ] 2020
+[ಕ][್ತ] -> [ಕ್ತ] 2021
+[ುದ][್ಧ] -> [ುದ್ಧ] 2022
+[ ए][स] -> [ एस] 2023
+[ p][os] -> [ pos] 2024
+[್ಯ][ದ] -> [್ಯದ] 2025
+[ent][ial] -> [ential] 2026
+[ ಯ][ೋಜ] -> [ ಯೋಜ] 2027
+[ा�][�] -> [ाब] 2028
+[ क][म] -> [ कम] 2029
+[t][ain] -> [tain] 2030
+[ H][er] -> [ Her] 2031
+[ se][em] -> [ seem] 2032
+[ sa][f] -> [ saf] 2033
+[ t][urn] -> [ turn] 2034
+[ re][al] -> [ real] 2035
+[on][y] -> [ony] 2036
+[ह][र] -> [हर] 2037
+[ he][re] -> [ here] 2038
+[ cont][in] -> [ contin] 2039
+[ur][y] -> [ury] 2040
+[ ][es] -> [ es] 2041
+[ m][em] -> [ mem] 2042
+[t][he] -> [the] 2043
+[ d][ig] -> [ dig] 2044
+[it][ive] -> [itive] 2045
+[b][s] -> [bs] 2046
+[ಿವ][ೆ] -> [ಿವೆ] 2047
+[ ಆದ][ರೆ] -> [ ಆದರೆ] 2048
+[ �][�] -> [ ओ] 2049
+['][t] -> ['t] 2050
+[ ][ide] -> [ ide] 2051
+[ e][t] -> [ et] 2052
+[ t][est] -> [ test] 2053
+[ c][ase] -> [ case] 2054
+[5][0] -> [50] 2055
+[ ಸ][ೇರ] -> [ ಸೇರ] 2056
+[ क][्ल] -> [ क्ल] 2057
+[�][�] -> [थ] 2058
+[ ಆ][ತ] -> [ ಆತ] 2059
+[ी][क] -> [ीक] 2060
+[ान][ी] -> [ानी] 2061
+[ g][row] -> [ grow] 2062
+[d][e] -> [de] 2063
+[ಿಸ][ುತ್ತದೆ] -> [ಿಸುತ್ತದೆ] 2064
+[ ब][्र] -> [ ब्र] 2065
+[ Americ][an] -> [ American] 2066
+[t][on] -> [ton] 2067
+[ ए][ल] -> [ एल] 2068
+[as][ter] -> [aster] 2069
+[ ap][pe] -> [ appe] 2070
+[aj][or] -> [ajor] 2071
+[िक][ा] -> [िका] 2072
+[ಾನ][್ಯ] -> [ಾನ್ಯ] 2073
+[ I][s] -> [ Is] 2074
+[ै][ंड] -> [ैंड] 2075
+[ಿಯ][ು] -> [ಿಯು] 2076
+[ जिस][में] -> [ जिसमें] 2077
+[ en][ergy] -> [ energy] 2078
+[ ][=] -> [ =] 2079
+[ರ್�][�] -> [ರ್ಸ] 2080
+[in][ess] -> [iness] 2081
+[ w][ar] -> [ war] 2082
+[ L][e] -> [ Le] 2083
+[a][x] -> [ax] 2084
+[ �][�] -> [ ‘] 2085
+[ P][l] -> [ Pl] 2086
+[ re][du] -> [ redu] 2087
+[ �][�] -> [ ಐ] 2088
+[ಂ][ಧ] -> [ಂಧ] 2089
+[ಿಯ][ಾ] -> [ಿಯಾ] 2090
+[ang][u] -> [angu] 2091
+[ೈ][ನ್] -> [ೈನ್] 2092
+[l][ine] -> [line] 2093
+[ S][h] -> [ Sh] 2094
+[ com][e] -> [ come] 2095
+[ ch][ange] -> [ change] 2096
+[as][on] -> [ason] 2097
+[ ob][ject] -> [ object] 2098
+[ಿಂ][ಗ್] -> [ಿಂಗ್] 2099
+[ೈ][ನ] -> [ೈನ] 2100
+[ ಸ][ು] -> [ ಸು] 2101
+[ಬ][ಹ] -> [ಬಹ] 2102
+[्�][�] -> [्ब] 2103
+[om][s] -> [oms] 2104
+[ म][ह] -> [ मह] 2105
+[er][c] -> [erc] 2106
+[an][k] -> [ank] 2107
+[1][8] -> [18] 2108
+[nt][il] -> [ntil] 2109
+[os][ed] -> [osed] 2110
+[ f][ood] -> [ food] 2111
+[ ಕ][ು] -> [ ಕು] 2112
+[e][x] -> [ex] 2113
+[ ][$] -> [ $] 2114
+[b][ers] -> [bers] 2115
+[ent][ion] -> [ention] 2116
+[ E][uro] -> [ Euro] 2117
+[ a][ir] -> [ air] 2118
+[ಸ][್ಟ] -> [ಸ್ಟ] 2119
+[ಿಕ][ೆ] -> [ಿಕೆ] 2120
+[ be][h] -> [ beh] 2121
+[क][ों] -> [कों] 2122
+[ವಾಗ][ಿದೆ] -> [ವಾಗಿದೆ] 2123
+[್ಯ][ಾನ] -> [್ಯಾನ] 2124
+[ि�][�] -> [िद] 2125
+[o][on] -> [oon] 2126
+[ le][t] -> [ let] 2127
+[ me][ans] -> [ means] 2128
+[l][ess] -> [less] 2129
+[ s][ince] -> [ since] 2130
+[ आ][य] -> [ आय] 2131
+[�][�] -> [ठ] 2132
+[�][�] -> [ः] 2133
+[ಬ][್ಬ] -> [ಬ್ಬ] 2134
+[ स][क] -> [ सक] 2135
+[ರ್�][�] -> [ರ್ಮ] 2136
+[त][ि] -> [ति] 2137
+[ कर][ता] -> [ करता] 2138
+[oc][i] -> [oci] 2139
+[ ವಿ�][�] -> [ ವಿಶ] 2140
+[ॉ][र्�] -> [ॉर्�] 2141
+[ n][ame] -> [ name] 2142
+[ b][re] -> [ bre] 2143
+[ L][ena] -> [ Lena] 2144
+[ಿತ][್ರ] -> [ಿತ್ರ] 2145
+[ात][्र] -> [ात्र] 2146
+[ अ][म] -> [ अम] 2147
+[ go][vern] -> [ govern] 2148
+[ 1][3] -> [ 13] 2149
+[े�][�] -> [ेष] 2150
+[i][qu] -> [iqu] 2151
+[ b][as] -> [ bas] 2152
+[ p][ut] -> [ put] 2153
+[ e][arly] -> [ early] 2154
+[ me][an] -> [ mean] 2155
+[n][ce] -> [nce] 2156
+[o][h] -> [oh] 2157
+[u][g] -> [ug] 2158
+[ं�][�] -> [ंज] 2159
+[h][or] -> [hor] 2160
+[ ha][pp] -> [ happ] 2161
+[iv][en] -> [iven] 2162
+[ dire][ct] -> [ direct] 2163
+[ ಅ][ಂತ] -> [ ಅಂತ] 2164
+[im][e] -> [ime] 2165
+[ं�][�] -> [ंध] 2166
+[ाड][़] -> [ाड़] 2167
+[re][nt] -> [rent] 2168
+[ o][per] -> [ oper] 2169
+[ go][ing] -> [ going] 2170
+[ sh][ow] -> [ show] 2171
+[ How][ever] -> [ However] 2172
+[ivid][ual] -> [ividual] 2173
+[cc][ess] -> [ccess] 2174
+[i][od] -> [iod] 2175
+[ g][u] -> [ gu] 2176
+[ cont][ro] -> [ contro] 2177
+[ 1][1] -> [ 11] 2178
+[ಪ][ಡ] -> [ಪಡ] 2179
+[ st][e] -> [ ste] 2180
+[ क][िए] -> [ किए] 2181
+[ m][ar] -> [ mar] 2182
+[iv][ed] -> [ived] 2183
+[ o][il] -> [ oil] 2184
+[our][ces] -> [ources] 2185
+[ಳ][ು] -> [ಳು] 2186
+[ ಉ][ತ್ತ] -> [ ಉತ್ತ] 2187
+[ip][s] -> [ips] 2188
+[ m][en] -> [ men] 2189
+[ N][av] -> [ Nav] 2190
+[ p][ur] -> [ pur] 2191
+[en][se] -> [ense] 2192
+[ ca][us] -> [ caus] 2193
+[ ][&] -> [ &] 2194
+[if][ied] -> [ified] 2195
+[ion][al] -> [ional] 2196
+[ u][ntil] -> [ until] 2197
+[ an][t] -> [ ant] 2198
+[ व][े] -> [ वे] 2199
+[ हु�][�] -> [ हुए] 2200
+[ar][ing] -> [aring] 2201
+[ा�][�] -> [ाण] 2202
+[ A][ll] -> [ All] 2203
+[id][ed] -> [ided] 2204
+[ m][et] -> [ met] 2205
+[ ke][ep] -> [ keep] 2206
+[ re][st] -> [ rest] 2207
+[al][k] -> [alk] 2208
+[ up][on] -> [ upon] 2209
+[ h][ard] -> [ hard] 2210
+[ A][nt] -> [ Ant] 2211
+[i][or] -> [ior] 2212
+[ m][at] -> [ mat] 2213
+[ rese][arch] -> [ research] 2214
+[ಾರ][ಂಭ] -> [ಾರಂಭ] 2215
+[್ಗಳ][ು] -> [್ಗಳು] 2216
+[ b][est] -> [ best] 2217
+[c][le] -> [cle] 2218
+[ie][ve] -> [ieve] 2219
+[ p][ap] -> [ pap] 2220
+[ೆ][ಂಬ] -> [ೆಂಬ] 2221
+[ೆ�][�] -> [ೆಮ] 2222
+[�][�] -> [ಅ] 2223
+[p][or] -> [por] 2224
+[w][are] -> [ware] 2225
+[c][ing] -> [cing] 2226
+[ो][म] -> [ोम] 2227
+[ bec][ome] -> [ become] 2228
+[ se][ver] -> [ sever] 2229
+[िय][न] -> [ियन] 2230
+[ater][ial] -> [aterial] 2231
+[क][्स] -> [क्स] 2232
+[ व][ह] -> [ वह] 2233
+[ t][re] -> [ tre] 2234
+[ e][y] -> [ ey] 2235
+[ W][ith] -> [ With] 2236
+[ includ][ing] -> [ including] 2237
+[ bl][ood] -> [ blood] 2238
+[ f][our] -> [ four] 2239
+[ಡ][ೆಯ] -> [ಡೆಯ] 2240
+[ಂತ][ೆ] -> [ಂತೆ] 2241
+[ or][der] -> [ order] 2242
+[ ][ident] -> [ ident] 2243
+[l][ing] -> [ling] 2244
+[ d][em] -> [ dem] 2245
+[स][र] -> [सर] 2246
+[�][़] -> [�़] 2247
+[ p][ar] -> [ par] 2248
+[ ल][ेक] -> [ लेक] 2249
+[ आ][ई] -> [ आई] 2250
+[ ne][ver] -> [ never] 2251
+[us][s] -> [uss] 2252
+[r][id] -> [rid] 2253
+[ೋ][ಟ] -> [ೋಟ] 2254
+[ he][ad] -> [ head] 2255
+[ो][न] -> [ोन] 2256
+[ ग][्र] -> [ ग्र] 2257
+[ उप][योग] -> [ उपयोग] 2258
+[ T][y] -> [ Ty] 2259
+[ect][ion] -> [ection] 2260
+[ c][ertain] -> [ certain] 2261
+[ R][ail] -> [ Rail] 2262
+[ ab][le] -> [ able] 2263
+[en][g] -> [eng] 2264
+[ ड][ी] -> [ डी] 2265
+[ वाल][े] -> [ वाले] 2266
+[ sy][m] -> [ sym] 2267
+[ ty][pe] -> [ type] 2268
+[ gen][es] -> [ genes] 2269
+[it][ions] -> [itions] 2270
+[ st][ate] -> [ state] 2271
+[ produ][ct] -> [ product] 2272
+[ m][ajor] -> [ major] 2273
+[ ][ess] -> [ ess] 2274
+[್�][�] -> [್ಜ] 2275
+[ಿ�][�] -> [ಿಜ] 2276
+[ th][ink] -> [ think] 2277
+[ उ][स] -> [ उस] 2278
+[ar][ning] -> [arning] 2279
+[ क][ु�] -> [ कु�] 2280
+[ e][l] -> [ el] 2281
+[ತ][್ರ] -> [ತ್ರ] 2282
+[ೀ][ಕ] -> [ೀಕ] 2283
+[ज][न] -> [जन] 2284
+[ ब][ार] -> [ बार] 2285
+[ M][r] -> [ Mr] 2286
+[un][e] -> [une] 2287
+[ स्क][ूल] -> [ स्कूल] 2288
+[ r][un] -> [ run] 2289
+[at][ic] -> [atic] 2290
+[ s][it] -> [ sit] 2291
+[ಿ�][�] -> [ಿಹ] 2292
+[ de][p] -> [ dep] 2293
+[ro][du] -> [rodu] 2294
+[ಕ][ೆ] -> [ಕೆ] 2295
+[िय][र] -> [ियर] 2296
+[ु][क्त] -> [ुक्त] 2297
+[ te][xt] -> [ text] 2298
+[ S][o] -> [ So] 2299
+[ ind][ividual] -> [ individual] 2300
+[ m][ark] -> [ mark] 2301
+[v][is] -> [vis] 2302
+[ಿದ್ದ][ರು] -> [ಿದ್ದರು] 2303
+[ ಕ][ೇ] -> [ ಕೇ] 2304
+[ स][ब] -> [ सब] 2305
+[ ल][ग] -> [ लग] 2306
+[ grou][p] -> [ group] 2307
+[ an][al] -> [ anal] 2308
+[ ल][ी] -> [ ली] 2309
+[े�][�] -> [ेख] 2310
+[ord][ing] -> [ording] 2311
+[ ret][urn] -> [ return] 2312
+[ N][ew] -> [ New] 2313
+[ soc][ial] -> [ social] 2314
+[oo][ks] -> [ooks] 2315
+[ ಕ][ಡ] -> [ ಕಡ] 2316
+[ಡೆ][ದ] -> [ಡೆದ] 2317
+[ 3][0] -> [ 30] 2318
+[ाइ][न] -> [ाइन] 2319
+[ l][ow] -> [ low] 2320
+[ sub][ject] -> [ subject] 2321
+[ द][े] -> [ दे] 2322
+[o][ad] -> [oad] 2323
+[ ए][म] -> [ एम] 2324
+[ ಬ][ೈ] -> [ ಬೈ] 2325
+[ develop][ment] -> [ development] 2326
+[ic][le] -> [icle] 2327
+[ c][ause] -> [ cause] 2328
+[if][y] -> [ify] 2329
+[ent][al] -> [ental] 2330
+[con][om] -> [conom] 2331
+[ partic][ular] -> [ particular] 2332
+[ sever][al] -> [ several] 2333
+[ c][reat] -> [ creat] 2334
+[ಗಳ][ಿಗೆ] -> [ಗಳಿಗೆ] 2335
+[ಗ][ೊಳ] -> [ಗೊಳ] 2336
+[ೊ][ಡ] -> [ೊಡ] 2337
+[ೆ][ಳ] -> [ೆಳ] 2338
+[ उन्ह][ें] -> [ उन्हें] 2339
+[f][ace] -> [face] 2340
+[eth][ing] -> [ething] 2341
+[iz][ation] -> [ization] 2342
+[ pre][vent] -> [ prevent] 2343
+[ಲ][ಾಗ] -> [ಲಾಗ] 2344
+[ೃ][ತ] -> [ೃತ] 2345
+[1][1] -> [11] 2346
+[ द][ौर] -> [ दौर] 2347
+[ man][ag] -> [ manag] 2348
+[ उन][के] -> [ उनके] 2349
+[h][ing] -> [hing] 2350
+[ a][way] -> [ away] 2351
+[ कर][ते] -> [ करते] 2352
+[ c][ash] -> [ cash] 2353
+[ N][arn] -> [ Narn] 2354
+[ includ][e] -> [ include] 2355
+[ s][ent] -> [ sent] 2356
+[ ಸ್ಥ][ಳ] -> [ ಸ್ಥಳ] 2357
+[ ಪ್ರ][ಾರಂಭ] -> [ ಪ್ರಾರಂಭ] 2358
+[ ಬ][್ರ] -> [ ಬ್ರ] 2359
+[angu][age] -> [anguage] 2360
+[ pro][per] -> [ proper] 2361
+[ ಪ್ರ][ಮ] -> [ ಪ್ರಮ] 2362
+[ 18][3] -> [ 183] 2363
+[u][se] -> [use] 2364
+[�][�] -> [—] 2365
+[ूर्�][��] -> [ूर्व] 2366
+[ि�][�] -> [िण] 2367
+[at][s] -> [ats] 2368
+[ab][ly] -> [ably] 2369
+[ other][s] -> [ others] 2370
+[ m][ove] -> [ move] 2371
+[ber][t] -> [bert] 2372
+[ f][ore] -> [ fore] 2373
+[ o][pp] -> [ opp] 2374
+[ qu][est] -> [ quest] 2375
+[ೇ][ಟ] -> [ೇಟ] 2376
+[ ಸಮ][ಯ] -> [ ಸಮಯ] 2377
+[ th][ings] -> [ things] 2378
+[ c][r] -> [ cr] 2379
+[vir][on] -> [viron] 2380
+[ b][ene] -> [ bene] 2381
+[ार्�][�] -> [ार्ट] 2382
+[ sec][ond] -> [ second] 2383
+[ inv][ol] -> [ invol] 2384
+[i][ans] -> [ians] 2385
+[ al][ways] -> [ always] 2386
+[m][an] -> [man] 2387
+[ b][us] -> [ bus] 2388
+[ ph][ys] -> [ phys] 2389
+[ tim][es] -> [ times] 2390
+[ 19][5] -> [ 195] 2391
+[ ज][न] -> [ जन] 2392
+[�][�़] -> [ढ़] 2393
+[ re][pl] -> [ repl] 2394
+[is][on] -> [ison] 2395
+[ स][ह] -> [ सह] 2396
+[y][l] -> [yl] 2397
+[ s][l] -> [ sl] 2398
+[ count][ry] -> [ country] 2399
+[ill][ion] -> [illion] 2400
+[ O][r] -> [ Or] 2401
+[ Euro][pe] -> [ Europe] 2402
+[ gener][al] -> [ general] 2403
+[ಿಗಳ][ನ್ನು] -> [ಿಗಳನ್ನು] 2404
+[am][es] -> [ames] 2405
+[ak][en] -> [aken] 2406
+[ro][p] -> [rop] 2407
+[ ne][xt] -> [ next] 2408
+[b][e] -> [be] 2409
+[c][er] -> [cer] 2410
+[ ಸ][್ಟ] -> [ ಸ್ಟ] 2411
+[ಬಹ][ುದು] -> [ಬಹುದು] 2412
+[ द][ू] -> [ दू] 2413
+[i][ents] -> [ients] 2414
+[ Com][m] -> [ Comm] 2415
+[ अ][ल] -> [ अल] 2416
+[�][�] -> [अ] 2417
+[ प्र][द] -> [ प्रद] 2418
+[eth][od] -> [ethod] 2419
+[or][th] -> [orth] 2420
+[ v][ol] -> [ vol] 2421
+[r][ation] -> [ration] 2422
+[ p][ri] -> [ pri] 2423
+[ ವ][ಿವ] -> [ ವಿವ] 2424
+[ೀ][ತ] -> [ೀತ] 2425
+[್�][�] -> [್ಹ] 2426
+[m][e] -> [me] 2427
+[ m][aterial] -> [ material] 2428
+[ s][ays] -> [ says] 2429
+[ e][le] -> [ ele] 2430
+[ कु�][�] -> [ कुछ] 2431
+[ ind][ust] -> [ indust] 2432
+[ p][ost] -> [ post] 2433
+[ಿನ][್] -> [ಿನ್] 2434
+[ श][्र] -> [ श्र] 2435
+[ le][arn] -> [ learn] 2436
+[ ris][k] -> [ risk] 2437
+[ै][न] -> [ैन] 2438
+[ फ][िल] -> [ फिल] 2439
+[v][ious] -> [vious] 2440
+[our][ce] -> [ource] 2441
+[ard][s] -> [ards] 2442
+[ imp][ro] -> [ impro] 2443
+[ st][at] -> [ stat] 2444
+[ಂತ][್ರ] -> [ಂತ್ರ] 2445
+[ರ][ೆಯ] -> [ರೆಯ] 2446
+[im][es] -> [imes] 2447
+[ लेक][िन] -> [ लेकिन] 2448
+[ d][one] -> [ done] 2449
+[im][ate] -> [imate] 2450
+[i][ver] -> [iver] 2451
+[ig][ure] -> [igure] 2452
+[ poss][ible] -> [ possible] 2453
+[ fam][ily] -> [ family] 2454
+[a][pe] -> [ape] 2455
+[ ಪೂರ್�][�] -> [ ಪೂರ್ವ] 2456
+[ಮ][್] -> [ಮ್] 2457
+[ र][ाष्ट्र] -> [ राष्ट्र] 2458
+[ु][त] -> [ुत] 2459
+[t][o] -> [to] 2460
+[ prof][ess] -> [ profess] 2461
+[viron][ment] -> [vironment] 2462
+[ ][Z] -> [ Z] 2463
+[ w][ind] -> [ wind] 2464
+[oc][us] -> [ocus] 2465
+[ul][ation] -> [ulation] 2466
+[ra][p] -> [rap] 2467
+[ ಅಧ][್ಯ] -> [ ಅಧ್ಯ] 2468
+[ಿ�][�] -> [ಿಧ] 2469
+[ ಇ][ತರ] -> [ ಇತರ] 2470
+[ಿಯ][ೋ] -> [ಿಯೋ] 2471
+[ wor][ds] -> [ words] 2472
+[em][ber] -> [ember] 2473
+[ O][ne] -> [ One] 2474
+[ ಟ][ಿ] -> [ ಟಿ] 2475
+[ 19][3] -> [ 193] 2476
+[್ಯ][ಾರ] -> [್ಯಾರ] 2477
+[ ಪ][ಂದ] -> [ ಪಂದ] 2478
+[at][ural] -> [atural] 2479
+[ is][s] -> [ iss] 2480
+[ C][an] -> [ Can] 2481
+[ ಹೊ][ಸ] -> [ ಹೊಸ] 2482
+[ा�][�] -> [ाफ] 2483
+[ pro][m] -> [ prom] 2484
+[ S][ome] -> [ Some] 2485
+[2][00] -> [200] 2486
+[A][D] -> [AD] 2487
+[ पुर][स्क] -> [ पुरस्क] 2488
+[ ह][ाल] -> [ हाल] 2489
+[ प्र][ाप्त] -> [ प्राप्त] 2490
+[t][t] -> [tt] 2491
+[ S][p] -> [ Sp] 2492
+[ som][ething] -> [ something] 2493
+[ A][fter] -> [ After] 2494
+[ A][d] -> [ Ad] 2495
+[al][ity] -> [ality] 2496
+[ spec][ial] -> [ special] 2497
+[ per][iod] -> [ period] 2498
+[ೇ][ಷ] -> [ೇಷ] 2499
+[ h][ouse] -> [ house] 2500
+[le][y] -> [ley] 2501
+[ to][p] -> [ top] 2502
+[ mon][th] -> [ month] 2503
+[ cour][se] -> [ course] 2504
+[ ಪ್ರದ][ೇಶ] -> [ ಪ್ರದೇಶ] 2505
+[ಾ�][�] -> [ಾಶ] 2506
+[ follow][ing] -> [ following] 2507
+[ स][ाम] -> [ साम] 2508
+[ to][get] -> [ toget] 2509
+[ toget][her] -> [ together] 2510
+[र्�][�] -> [र्स] 2511
+[ c][ells] -> [ cells] 2512
+[ पुरस्क][ार] -> [ पुरस्कार] 2513
+[ क][ई] -> [ कई] 2514
+[i][ot] -> [iot] 2515
+[ bene][f] -> [ benef] 2516
+[ an][im] -> [ anim] 2517
+[ hist][ory] -> [ history] 2518
+[are][nt] -> [arent] 2519
+[ Un][iversity] -> [ University] 2520
+[ ಎರಡ][ು] -> [ ಎರಡು] 2521
+[ೂ][ಪ] -> [ೂಪ] 2522
+[ व][िक] -> [ विक] 2523
+[ी][ल] -> [ील] 2524
+[ निर्�][�] -> [ निर्म] 2525
+[ಜ][ಾಂಟ] -> [ಜಾಂಟ] 2526
+[e][g] -> [eg] 2527
+[ E][arth] -> [ Earth] 2528
+[ dis][e] -> [ dise] 2529
+[್ಡ][್] -> [್ಡ್] 2530
+[ಾಲ][ಯ] -> [ಾಲಯ] 2531
+[ೀ][ನ] -> [ೀನ] 2532
+[ ನ][ಿಯ] -> [ ನಿಯ] 2533
+[ अम][ेर] -> [ अमेर] 2534
+[ l][ine] -> [ line] 2535
+[ w][r] -> [ wr] 2536
+[l][ish] -> [lish] 2537
+[ar][c] -> [arc] 2538
+[2][1] -> [21] 2539
+[ ಬೈ][ಜಾಂಟ] -> [ ಬೈಜಾಂಟ] 2540
+[ �][�] -> [ —] 2541
+[ic][ult] -> [icult] 2542
+[ R][es] -> [ Res] 2543
+[ wh][y] -> [ why] 2544
+[ ab][ove] -> [ above] 2545
+[ all][ow] -> [ allow] 2546
+[ d][ays] -> [ days] 2547
+[iel][d] -> [ield] 2548
+[ ಮೂಲ][ಕ] -> [ ಮೂಲಕ] 2549
+[ pro][ject] -> [ project] 2550
+[ am][ong] -> [ among] 2551
+[n][er] -> [ner] 2552
+[ or][ig] -> [ orig] 2553
+[ too][k] -> [ took] 2554
+[ n][ight] -> [ night] 2555
+[्�][�] -> [्द] 2556
+[ क][ह] -> [ कह] 2557
+[ ಅದ][ರ] -> [ ಅದರ] 2558
+[ s][w] -> [ sw] 2559
+[ress][ion] -> [ression] 2560
+[ fe][el] -> [ feel] 2561
+[ ][Q] -> [ Q] 2562
+[ me][as] -> [ meas] 2563
+[ to][ol] -> [ tool] 2564
+[in][ation] -> [ination] 2565
+[o][pe] -> [ope] 2566
+[ಫ][್] -> [ಫ್] 2567
+[�][�] -> [ಃ] 2568
+[ಕ್ರ][ಮ] -> [ಕ್ರಮ] 2569
+[ ಸ][್ಪ] -> [ ಸ್ಪ] 2570
+[ro][gen] -> [rogen] 2571
+[ c][ame] -> [ came] 2572
+[ ज][िन] -> [ जिन] 2573
+[ Ant][ony] -> [ Antony] 2574
+[ le][ft] -> [ left] 2575
+[ 2][5] -> [ 25] 2576
+[ su][ccess] -> [ success] 2577
+[ಸ][್ಯ] -> [ಸ್ಯ] 2578
+[ क्ल][ब] -> [ क्लब] 2579
+[ eng][ine] -> [ engine] 2580
+[ ह][ी] -> [ ही] 2581
+[ent][ly] -> [ently] 2582
+[i][k] -> [ik] 2583
+[her][n] -> [hern] 2584
+[ le][arning] -> [ learning] 2585
+[ರ್ಶ][ನ] -> [ರ್ಶನ] 2586
+[ C][ent] -> [ Cent] 2587
+[ d][el] -> [ del] 2588
+[ ha][ving] -> [ having] 2589
+[v][an] -> [van] 2590
+[ in][c] -> [ inc] 2591
+[ಿ�][�] -> [ಿಳ] 2592
+[ut][ure] -> [uture] 2593
+[ pl][ant] -> [ plant] 2594
+[c][rib] -> [crib] 2595
+[ंब][र] -> [ंबर] 2596
+[ो][क] -> [ोक] 2597
+[ं�][�] -> [ंख] 2598
+[th][ing] -> [thing] 2599
+[ spec][ies] -> [ species] 2600
+[ Rail][road] -> [ Railroad] 2601
+[ f][ar] -> [ far] 2602
+[ h][old] -> [ hold] 2603
+[ beg][in] -> [ begin] 2604
+[ det][erm] -> [ determ] 2605
+[ ಅ][ಡ] -> [ ಅಡ] 2606
+[ ಪಂದ][್ಯ] -> [ ಪಂದ್ಯ] 2607
+[ m][aking] -> [ making] 2608
+[ टी][म] -> [ टीम] 2609
+[ p][ot] -> [ pot] 2610
+[ರ][ಿ] -> [ರಿ] 2611
+[ ज][ै] -> [ जै] 2612
+[्य][ालय] -> [्यालय] 2613
+[ प][ी] -> [ पी] 2614
+[ अप][नी] -> [ अपनी] 2615
+[ �][�] -> [ ध] 2616
+[ am][ount] -> [ amount] 2617
+[r][an] -> [ran] 2618
+[ aut][hor] -> [ author] 2619
+[ treat][ment] -> [ treatment] 2620
+[ sc][ient] -> [ scient] 2621
+[ಣ][ೆ] -> [ಣೆ] 2622
+[ವ][ರೆ] -> [ವರೆ] 2623
+[ಾರ][್ಕ] -> [ಾರ್ಕ] 2624
+[ सब][से] -> [ सबसे] 2625
+[ अन][्य] -> [ अन्य] 2626
+[्र][ै] -> [्रै] 2627
+[ell][a] -> [ella] 2628
+[ f][ocus] -> [ focus] 2629
+[pt][ion] -> [ption] 2630
+[ again][st] -> [ against] 2631
+[ wom][en] -> [ women] 2632
+[ಾಷ][್ಟ್ರ] -> [ಾಷ್ಟ್ರ] 2633
+[ ಸಂ][ಪ] -> [ ಸಂಪ] 2634
+[ि�][�] -> [िध] 2635
+[्र][ी] -> [्री] 2636
+[�][�] -> [आ] 2637
+[s][w] -> [sw] 2638
+[ ಬಳ][ಸ] -> [ ಬಳಸ] 2639
+[ se][en] -> [ seen] 2640
+[म][ान] -> [मान] 2641
+[ s][n] -> [ sn] 2642
+[ f][in] -> [ fin] 2643
+[ al][ong] -> [ along] 2644
+[it][her] -> [ither] 2645
+[pe][cial] -> [pecial] 2646
+[ೇ][ಜ] -> [ೇಜ] 2647
+[ ಸಾಮ][ಾನ್ಯ] -> [ ಸಾಮಾನ್ಯ] 2648
+[ಲ][್ಪ] -> [ಲ್ಪ] 2649
+[ �][�] -> [ ಧ] 2650
+[ಿದ್ದ][ಾರೆ] -> [ಿದ್ದಾರೆ] 2651
+[ c][ult] -> [ cult] 2652
+[ l][ist] -> [ list] 2653
+[ರ್�][�] -> [ರ್ಟ] 2654
+[ಿಕ][ಾರ] -> [ಿಕಾರ] 2655
+[ E][n] -> [ En] 2656
+[ic][ro] -> [icro] 2657
+[ श][हर] -> [ शहर] 2658
+[ उत][्प] -> [ उत्प] 2659
+[ d][am] -> [ dam] 2660
+[ इस][े] -> [ इसे] 2661
+[ ಹೊ][ರ] -> [ ಹೊರ] 2662
+[ಲ][ೆ] -> [ಲೆ] 2663
+[ಾ�][�] -> [ಾಬ] 2664
+[ ज][ारी] -> [ जारी] 2665
+[र्�][�] -> [र्ट] 2666
+[ p][ract] -> [ pract] 2667
+[um][ent] -> [ument] 2668
+[ t][ell] -> [ tell] 2669
+[ 18][6] -> [ 186] 2670
+[र्�][�] -> [र्म] 2671
+[ sh][are] -> [ share] 2672
+[ m][ethod] -> [ method] 2673
+[ chang][es] -> [ changes] 2674
+[ t][ry] -> [ try] 2675
+[ ass][oci] -> [ associ] 2676
+[ �][�] -> [ ಘ] 2677
+[ ನಡ][ುವ] -> [ ನಡುವ] 2678
+[ prog][ram] -> [ program] 2679
+[त][र] -> [तर] 2680
+[ add][ition] -> [ addition] 2681
+[ c][ut] -> [ cut] 2682
+[ G][e] -> [ Ge] 2683
+[ र][ख] -> [ रख] 2684
+[er][al] -> [eral] 2685
+[िय][म] -> [ियम] 2686
+[ि�][�] -> [िख] 2687
+[ Her][bert] -> [ Herbert] 2688
+[om][m] -> [omm] 2689
+[a][im] -> [aim] 2690
+[ to][day] -> [ today] 2691
+[ acc][ess] -> [ access] 2692
+[ educ][ation] -> [ education] 2693
+[ re][ce] -> [ rece] 2694
+[ಿಸ][ಿ] -> [ಿಸಿ] 2695
+[ಿಕ][ೆಯ] -> [ಿಕೆಯ] 2696
+[ Wh][ile] -> [ While] 2697
+[ occ][ur] -> [ occur] 2698
+[en][n] -> [enn] 2699
+[ we][e] -> [ wee] 2700
+[े�][�] -> [ेद] 2701
+[ೀವ][ನ] -> [ೀವನ] 2702
+[ಿಸಲ][ಾಗ] -> [ಿಸಲಾಗ] 2703
+[ಬ][್ಲ] -> [ಬ್ಲ] 2704
+[ क][्र] -> [ क्र] 2705
+[ sk][ills] -> [ skills] 2706
+[ l][anguage] -> [ language] 2707
+[ f][ac] -> [ fac] 2708
+[i][um] -> [ium] 2709
+[ char][acter] -> [ character] 2710
+[ म][िल] -> [ मिल] 2711
+[िक][्ष] -> [िक्ष] 2712
+[p][ar] -> [par] 2713
+[th][ough] -> [though] 2714
+[ B][y] -> [ By] 2715
+[ bec][ame] -> [ became] 2716
+[ a][ge] -> [ age] 2717
+[ ಒಳ][ಗೊಂಡ] -> [ ಒಳಗೊಂಡ] 2718
+[ ಅ][ಭ] -> [ ಅಭ] 2719
+[स][्थ] -> [स्थ] 2720
+[te][nd] -> [tend] 2721
+[o][ver] -> [over] 2722
+[ S][y] -> [ Sy] 2723
+[ G][erm] -> [ Germ] 2724
+[ward][s] -> [wards] 2725
+[ir][d] -> [ird] 2726
+[ प्र][श] -> [ प्रश] 2727
+[ diff][icult] -> [ difficult] 2728
+[ g][iven] -> [ given] 2729
+[a][z] -> [az] 2730
+[is][ion] -> [ision] 2731
+[ ಕ][ಾಲ] -> [ ಕಾಲ] 2732
+[ತ][ೆ] -> [ತೆ] 2733
+[en][cy] -> [ency] 2734
+[ಿಂ][ಗ] -> [ಿಂಗ] 2735
+[�][�] -> [ष] 2736
+[ं][क] -> [ंक] 2737
+[ b][i] -> [ bi] 2738
+[ Un][ited] -> [ United] 2739
+[ comp][ut] -> [ comput] 2740
+[e][c] -> [ec] 2741
+[v][iew] -> [view] 2742
+[ st][ruct] -> [ struct] 2743
+[gg][est] -> [ggest] 2744
+[ contro][l] -> [ control] 2745
+[ I][I] -> [ II] 2746
+[ರ್�][�] -> [ರ್ಥ] 2747
+[ ex][erc] -> [ exerc] 2748
+[g][ar] -> [gar] 2749
+[ c][ost] -> [ cost] 2750
+[ib][ility] -> [ibility] 2751
+[ k][ind] -> [ kind] 2752
+[र्�][�] -> [र्श] 2753
+[og][n] -> [ogn] 2754
+[ de][v] -> [ dev] 2755
+[ri][end] -> [riend] 2756
+[ Ty][pe] -> [ Type] 2757
+[ f][re] -> [ fre] 2758
+[क्र][म] -> [क्रम] 2759
+[ हो][ने] -> [ होने] 2760
+[ t][ells] -> [ tells] 2761
+[ cur][rent] -> [ current] 2762
+[re][am] -> [ream] 2763
+[il][es] -> [iles] 2764
+[ B][rit] -> [ Brit] 2765
+[ ph][ot] -> [ phot] 2766
+[ಿಸ][ಿಕೊಂಡ] -> [ಿಸಿಕೊಂಡ] 2767
+[ಿನ][ಲ್ಲಿ] -> [ಿನಲ್ಲಿ] 2768
+[ f][uture] -> [ future] 2769
+[ p][ress] -> [ press] 2770
+[o][id] -> [oid] 2771
+[ b][at] -> [ bat] 2772
+[led][ge] -> [ledge] 2773
+[ खिल][ाड़] -> [ खिलाड़] 2774
+[re][g] -> [reg] 2775
+[ consid][er] -> [ consider] 2776
+[ a][ffect] -> [ affect] 2777
+[ sp][ace] -> [ space] 2778
+[ inf][l] -> [ infl] 2779
+[್ದ][ೇಶ] -> [್ದೇಶ] 2780
+[್�][�] -> [್ಣ] 2781
+[ು][ಡ] -> [ುಡ] 2782
+[ wor][ks] -> [ works] 2783
+[)][:] -> [):] 2784
+[िक][ी] -> [िकी] 2785
+[ort][un] -> [ortun] 2786
+[in][c] -> [inc] 2787
+[ su][re] -> [ sure] 2788
+[ govern][ment] -> [ government] 2789
+[ಿಕ][ೊಳ್ಳ] -> [ಿಕೊಳ್ಳ] 2790
+[s][ide] -> [side] 2791
+[op][s] -> [ops] 2792
+[pl][oy] -> [ploy] 2793
+[ ಒ][ಪ್ಪ] -> [ ಒಪ್ಪ] 2794
+[ f][ire] -> [ fire] 2795
+[ उ][द] -> [ उद] 2796
+[ prov][ide] -> [ provide] 2797
+[ ex][ist] -> [ exist] 2798
+[ str][ong] -> [ strong] 2799
+[om][es] -> [omes] 2800
+[ level][s] -> [ levels] 2801
+[ d][om] -> [ dom] 2802
+[ W][or] -> [ Wor] 2803
+[or][ies] -> [ories] 2804
+[ mod][el] -> [ model] 2805
+[ are][as] -> [ areas] 2806
+[ಾಕ][್] -> [ಾಕ್] 2807
+[ ಪ್ರ][ಯ] -> [ ಪ್ರಯ] 2808
+[ಿಂ][ತ] -> [ಿಂತ] 2809
+[ಾ�][�] -> [ಾಳ] 2810
+[ d][r] -> [ dr] 2811
+[ಾರ][ರ] -> [ಾರರ] 2812
+[ ಸ][ುರ] -> [ ಸುರ] 2813
+[1][3] -> [13] 2814
+[ सं�][�] -> [ संय] 2815
+[ु�][�] -> [ुआ] 2816
+[w][h] -> [wh] 2817
+[ av][ail] -> [ avail] 2818
+[ en][ough] -> [ enough] 2819
+[ag][n] -> [agn] 2820
+[ s][old] -> [ sold] 2821
+[ f][at] -> [ fat] 2822
+[ su][ggest] -> [ suggest] 2823
+[ ಇದ][ನ್ನು] -> [ ಇದನ್ನು] 2824
+[ೋ][ನ] -> [ೋನ] 2825
+[ೋ][ಡ] -> [ೋಡ] 2826
+[್ಯ][ಾಸ] -> [್ಯಾಸ] 2827
+[�][�] -> [इ] 2828
+[1][4] -> [14] 2829
+[ wor][d] -> [ word] 2830
+[ು][ಖ] -> [ುಖ] 2831
+[ h][ig] -> [ hig] 2832
+[ि�][�] -> [िड] 2833
+[ व][ी] -> [ वी] 2834
+[ C][le] -> [ Cle] 2835
+[ot][s] -> [ots] 2836
+[ ne][cess] -> [ necess] 2837
+[olog][ical] -> [ological] 2838
+[e][ad] -> [ead] 2839
+[ though][t] -> [ thought] 2840
+[ W][ar] -> [ War] 2841
+[ imp][act] -> [ impact] 2842
+[ ಪ][ಾಲ] -> [ ಪಾಲ] 2843
+[ दौर][ान] -> [ दौरान] 2844
+[ें][ट] -> [ेंट] 2845
+[र][ो] -> [रो] 2846
+[ you][ng] -> [ young] 2847
+[ಬ][ಂಧ] -> [ಬಂಧ] 2848
+[ fun][ction] -> [ function] 2849
+[र्�][�] -> [र्ज] 2850
+[ be][low] -> [ below] 2851
+[ w][ent] -> [ went] 2852
+[ain][ed] -> [ained] 2853
+[ inv][est] -> [ invest] 2854
+[1][5] -> [15] 2855
+[ rep][res] -> [ repres] 2856
+[ol][ution] -> [olution] 2857
+[ sh][ort] -> [ short] 2858
+[sel][ves] -> [selves] 2859
+[ re][ason] -> [ reason] 2860
+[ ವ][ಿದ] -> [ ವಿದ] 2861
+[ ��ुर][ू] -> [ शुरू] 2862
+[ j][oint] -> [ joint] 2863
+[ C][al] -> [ Cal] 2864
+[id][d] -> [idd] 2865
+[z][e] -> [ze] 2866
+[ರ][ಿಯ] -> [ರಿಯ] 2867
+[ how][ever] -> [ however] 2868
+[ t][aken] -> [ taken] 2869
+[ उत्प][ाद] -> [ उत्पाद] 2870
+[ sur][face] -> [ surface] 2871
+[ ne][g] -> [ neg] 2872
+[ l][ive] -> [ live] 2873
+[ ne][ar] -> [ near] 2874
+[ d][ra] -> [ dra] 2875
+[l][ished] -> [lished] 2876
+[ ra][d] -> [ rad] 2877
+[os][p] -> [osp] 2878
+[il][ar] -> [ilar] 2879
+[ ಹ][ೆಸ] -> [ ಹೆಸ] 2880
+[ प][ूर्व] -> [ पूर्व] 2881
+[ विश][्व] -> [ विश्व] 2882
+[2][2] -> [22] 2883
+[ c][ir] -> [ cir] 2884
+[ per][cent] -> [ percent] 2885
+[ b][ra] -> [ bra] 2886
+[र्�][�] -> [र्त] 2887
+[्य][म] -> [्यम] 2888
+[ आ][व] -> [ आव] 2889
+[्�][�] -> [्ड] 2890
+[ॉ][न] -> [ॉन] 2891
+[em][m] -> [emm] 2892
+[ं][स] -> [ंस] 2893
+[id][ence] -> [idence] 2894
+[um][e] -> [ume] 2895
+[ incre][ase] -> [ increase] 2896
+[is][ing] -> [ising] 2897
+[ ar][m] -> [ arm] 2898
+[ St][ates] -> [ States] 2899
+[ಲ][ವ] -> [ಲವ] 2900
+[್�][�] -> [್ಫ] 2901
+[ಿಮ][ೆ] -> [ಿಮೆ] 2902
+[ೋ][ಸ] -> [ೋಸ] 2903
+[र्�][�] -> [र्थ] 2904
+[on][es] -> [ones] 2905
+[ ಸ][ರಣ] -> [ ಸರಣ] 2906
+[ಮ][ಾನ] -> [ಮಾನ] 2907
+[ि�][�] -> [िह] 2908
+[ f][oot] -> [ foot] 2909
+[ ][ill] -> [ ill] 2910
+[m][ost] -> [most] 2911
+[ಗ][್ರ] -> [ಗ್ರ] 2912
+[ whe][ther] -> [ whether] 2913
+[ ಡ][ಿ] -> [ ಡಿ] 2914
+[ದ][್ಧ] -> [ದ್ಧ] 2915
+[ सम][य] -> [ समय] 2916
+[र्�][�] -> [र्ष] 2917
+[ m][ach] -> [ mach] 2918
+[ like][ly] -> [ likely] 2919
+[ Com][p] -> [ Comp] 2920
+[ೇ][ಖ] -> [ೇಖ] 2921
+[ग][ा] -> [गा] 2922
+[ंद][्र] -> [ंद्र] 2923
+[ en][vironment] -> [ environment] 2924
+[g][ed] -> [ged] 2925
+[ हु�][�] -> [ हुआ] 2926
+[ Narn][ia] -> [ Narnia] 2927
+[ de][ath] -> [ death] 2928
+[ M][ed] -> [ Med] 2929
+[ ಎ][ಂ] -> [ ಎಂ] 2930
+[ 200][8] -> [ 2008] 2931
+[ರ್�][�] -> [ರ್ಡ] 2932
+[ T][e] -> [ Te] 2933
+[il][ities] -> [ilities] 2934
+[oo][king] -> [ooking] 2935
+[ k][ids] -> [ kids] 2936
+[ s][ound] -> [ sound] 2937
+[ sa][w] -> [ saw] 2938
+[ l][ater] -> [ later] 2939
+[ d][est] -> [ dest] 2940
+[ system][s] -> [ systems] 2941
+[il][t] -> [ilt] 2942
+[is][ed] -> [ised] 2943
+[ bus][iness] -> [ business] 2944
+[ ಪ್ರ][ವ] -> [ ಪ್ರವ] 2945
+[ ಮು][ಖ್ಯ] -> [ ಮುಖ್ಯ] 2946
+[ ರ][ೈ] -> [ ರೈ] 2947
+[ ಸಮಯ][ದಲ್ಲಿ] -> [ ಸಮಯದಲ್ಲಿ] 2948
+[ h][ands] -> [ hands] 2949
+[ ಹ][ೇಳ] -> [ ಹೇಳ] 2950
+[ pap][er] -> [ paper] 2951
+[ಿಸ][್] -> [ಿಸ್] 2952
+[ l][aw] -> [ law] 2953
+[ proble][m] -> [ problem] 2954
+[ d][ri] -> [ dri] 2955
+[ore][d] -> [ored] 2956
+[ v][iew] -> [ view] 2957
+[ತ][ಿಹ] -> [ತಿಹ] 2958
+[ತಿಹ][ಾಸ] -> [ತಿಹಾಸ] 2959
+[ 200][7] -> [ 2007] 2960
+[am][p] -> [amp] 2961
+[ D][r] -> [ Dr] 2962
+[ ವ][ೈ] -> [ ವೈ] 2963
+[ cle][an] -> [ clean] 2964
+[ic][ation] -> [ication] 2965
+[ app][lic] -> [ applic] 2966
+[ ][ver] -> [ ver] 2967
+[ರ್�][�] -> [ರ್ಚ] 2968
+[ Th][at] -> [ That] 2969
+[as][ing] -> [asing] 2970
+[ ty][pes] -> [ types] 2971
+[ cent][ury] -> [ century] 2972
+[pecial][ly] -> [pecially] 2973
+[ ಆ][ಯ] -> [ ಆಯ] 2974
+[ು][ಲ] -> [ುಲ] 2975
+[ ಯ][ಾವ] -> [ ಯಾವ] 2976
+[et][y] -> [ety] 2977
+[ le][ast] -> [ least] 2978
+[ ಸಂ][ಬಂಧ] -> [ ಸಂಬಂಧ] 2979
+[ s][ing] -> [ sing] 2980
+[ v][i] -> [ vi] 2981
+[ ra][il] -> [ rail] 2982
+[ avail][able] -> [ available] 2983
+[I][V] -> [IV] 2984
+[ ಸುರ][ಂಗ] -> [ ಸುರಂಗ] 2985
+[ w][at] -> [ wat] 2986
+[ sign][ific] -> [ signific] 2987
+[ w][ays] -> [ ways] 2988
+[ ಯ][ು] -> [ ಯು] 2989
+[್ಯಾಂ][ಡ್] -> [್ಯಾಂಡ್] 2990
+[ su][per] -> [ super] 2991
+[ul][es] -> [ules] 2992
+[o][ice] -> [oice] 2993
+[y][n] -> [yn] 2994
+[1][6] -> [16] 2995
+[ो][ज] -> [ोज] 2996
+[m][ber] -> [mber] 2997
+[ru][ction] -> [ruction] 2998
+[ cre][ate] -> [ create] 2999
+[ us][es] -> [ uses] 3000
+[ g][ive] -> [ give] 3001
+[ proble][ms] -> [ problems] 3002
+[u][ed] -> [ued] 3003
+[ದ][ು] -> [ದು] 3004
+[ ಕ][್ಲ] -> [ ಕ್ಲ] 3005
+[್ವ][ಹ] -> [್ವಹ] 3006
+[ ಉತ್ತ][ರ] -> [ ಉತ್ತರ] 3007
+[ಾಗಿದ್ದ][ರು] -> [ಾಗಿದ್ದರು] 3008
+[्�][�] -> [्ञ] 3009
+[ an][sw] -> [ answ] 3010
+[ म][ूल] -> [ मूल] 3011
+[ ex][pect] -> [ expect] 3012
+[ಂ][ಚ] -> [ಂಚ] 3013
+[ स्थ][ित] -> [ स्थित] 3014
+[an][ces] -> [ances] 3015
+[e][b] -> [eb] 3016
+[ो][व] -> [ोव] 3017
+[ इस][के] -> [ इसके] 3018
+[er][y] -> [ery] 3019
+[ure][d] -> [ured] 3020
+[ M][iss] -> [ Miss] 3021
+[al][t] -> [alt] 3022
+[ab][ility] -> [ability] 3023
+[ off][ic] -> [ offic] 3024
+[t][al] -> [tal] 3025
+[ dig][est] -> [ digest] 3026
+[ l][ay] -> [ lay] 3027
+[ ಅ][ಪ] -> [ ಅಪ] 3028
+[ ಆ][ಡ] -> [ ಆಡ] 3029
+[ೋ][ದ] -> [ೋದ] 3030
+[ ವ][ಿರ] -> [ ವಿರ] 3031
+[ु�][�] -> [ुट] 3032
+[ n][atural] -> [ natural] 3033
+[ t][ru] -> [ tru] 3034
+[ S][oc] -> [ Soc] 3035
+[ ज][िल] -> [ जिल] 3036
+[ कर][ना] -> [ करना] 3037
+[ M][ay] -> [ May] 3038
+[os][is] -> [osis] 3039
+[ e][conom] -> [ econom] 3040
+[ col][le] -> [ colle] 3041
+[ ಸ][ಿ] -> [ ಸಿ] 3042
+[ ಕಡ][ಿಮೆ] -> [ ಕಡಿಮೆ] 3043
+[ ಮು][ಂದ] -> [ ಮುಂದ] 3044
+[ಾನ][ೆ] -> [ಾನೆ] 3045
+[ರ][ೂ] -> [ರೂ] 3046
+[ब][ॉल] -> [बॉल] 3047
+[क्ष][िण] -> [क्षिण] 3048
+[ m][akes] -> [ makes] 3049
+[at][or] -> [ator] 3050
+[ 1][00] -> [ 100] 3051
+[ ರ][ಲ್ಲಿ] -> [ ರಲ್ಲಿ] 3052
+[ writ][ing] -> [ writing] 3053
+[ spec][ific] -> [ specific] 3054
+[ap][s] -> [aps] 3055
+[ �][�] -> [ ई] 3056
+[ G][en] -> [ Gen] 3057
+[ap][an] -> [apan] 3058
+[w][w] -> [ww] 3059
+[ b][ased] -> [ based] 3060
+[ exper][ience] -> [ experience] 3061
+[ produ][ce] -> [ produce] 3062
+[ಾಷ್ಟ್ರ][ೀಯ] -> [ಾಷ್ಟ್ರೀಯ] 3063
+[ು][ತ] -> [ುತ] 3064
+[ೀ][ಟ] -> [ೀಟ] 3065
+[�][�] -> [ಆ] 3066
+[ pro][b] -> [ prob] 3067
+[ s][ide] -> [ side] 3068
+[ ल][ोग] -> [ लोग] 3069
+[ t][er] -> [ ter] 3070
+[in][a] -> [ina] 3071
+[ Americ][a] -> [ America] 3072
+[ g][reen] -> [ green] 3073
+[ा�][�] -> [ाश] 3074
+[े�][�] -> [ेय] 3075
+[ P][ip] -> [ Pip] 3076
+[ ][er] -> [ er] 3077
+[ f][ull] -> [ full] 3078
+[ wor][king] -> [ working] 3079
+[ soc][iety] -> [ society] 3080
+[rib][ut] -> [ribut] 3081
+[our][n] -> [ourn] 3082
+[ct][or] -> [ctor] 3083
+[ov][ed] -> [oved] 3084
+[ ನ][ಗ] -> [ ನಗ] 3085
+[ಲ್ಲ][ಿನ] -> [ಲ್ಲಿನ] 3086
+[ ಆ][ಟ] -> [ ಆಟ] 3087
+[ೀ][ಸ] -> [ೀಸ] 3088
+[ prot][ect] -> [ protect] 3089
+[et][imes] -> [etimes] 3090
+[ು][ಗಳನ್ನು] -> [ುಗಳನ್ನು] 3091
+[ प][ू] -> [ पू] 3092
+[ स][ेव] -> [ सेव] 3093
+[ M][ar] -> [ Mar] 3094
+[ा�][�] -> [ाट] 3095
+[ पर][िव] -> [ परिव] 3096
+[ st][ep] -> [ step] 3097
+[m][it] -> [mit] 3098
+[ro][wn] -> [rown] 3099
+[ us][ually] -> [ usually] 3100
+[rough][t] -> [rought] 3101
+[ on][ce] -> [ once] 3102
+[ob][al] -> [obal] 3103
+[ ಹೆಚ್ಚ][ು] -> [ ಹೆಚ್ಚು] 3104
+[ರ][್ನ] -> [ರ್ನ] 3105
+[ ಪ][ಡೆದ] -> [ ಪಡೆದ] 3106
+[ವರೆ][ಗೆ] -> [ವರೆಗೆ] 3107
+[ activ][ities] -> [ activities] 3108
+[ m][ind] -> [ mind] 3109
+[ಬ][ೇಕ] -> [ಬೇಕ] 3110
+[ 2][4] -> [ 24] 3111
+[ comp][any] -> [ company] 3112
+[ाल][ी] -> [ाली] 3113
+[ म][ुख] -> [ मुख] 3114
+[ाद][ी] -> [ादी] 3115
+[ स][र] -> [ सर] 3116
+[ J][apan] -> [ Japan] 3117
+[ur][ther] -> [urther] 3118
+[l][in] -> [lin] 3119
+[ab][or] -> [abor] 3120
+[ ನ][್ಯೂ] -> [ ನ್ಯೂ] 3121
+[ू][न] -> [ून] 3122
+[ techn][ology] -> [ technology] 3123
+[ s][at] -> [ sat] 3124
+[ಿ�][�] -> [ಿಚ] 3125
+[ु�][�] -> [ुब] 3126
+[el][f] -> [elf] 3127
+[क][्ट] -> [क्ट] 3128
+[ m][ot] -> [ mot] 3129
+[ व्य][क्त] -> [ व्यक्त] 3130
+[ur][ch] -> [urch] 3131
+[ M][AD] -> [ MAD] 3132
+[ MAD][S] -> [ MADS] 3133
+[ count][ries] -> [ countries] 3134
+[I][S] -> [IS] 3135
+[ g][re] -> [ gre] 3136
+[ ob][s] -> [ obs] 3137
+[ab][les] -> [ables] 3138
+[ಿ�][�] -> [ಿಬ] 3139
+[ಿನ][್ನ] -> [ಿನ್ನ] 3140
+[ �][�] -> [ ಓ] 3141
+[ te][ac] -> [ teac] 3142
+[ de][pend] -> [ depend] 3143
+[ rese][arc] -> [ researc] 3144
+[ con][c] -> [ conc] 3145
+[್ರ][ಾಜ] -> [್ರಾಜ] 3146
+[ 19][0] -> [ 190] 3147
+[ fl][ag] -> [ flag] 3148
+[ h][our] -> [ hour] 3149
+[ con][d] -> [ cond] 3150
+[u][le] -> [ule] 3151
+[ them][selves] -> [ themselves] 3152
+[ l][at] -> [ lat] 3153
+[ dise][ase] -> [ disease] 3154
+[ जिस][े] -> [ जिसे] 3155
+[ ವ್ಯ][ವ] -> [ ವ್ಯವ] 3156
+[ च][ै] -> [ चै] 3157
+[व][ल] -> [वल] 3158
+[ ब][ड़] -> [ बड़] 3159
+[ive][ly] -> [ively] 3160
+[ರ][ಿಂದ] -> [ರಿಂದ] 3161
+[ फ][िर] -> [ फिर] 3162
+[ sh][ip] -> [ ship] 3163
+[ के][ंद्र] -> [ केंद्र] 3164
+[ sym][pt] -> [ sympt] 3165
+[ co][ol] -> [ cool] 3166
+[al][f] -> [alf] 3167
+[र्�][�] -> [र्न] 3168
+[ क][ाम] -> [ काम] 3169
+[ 200][6] -> [ 2006] 3170
+[g][ers] -> [gers] 3171
+[ him][self] -> [ himself] 3172
+[od][e] -> [ode] 3173
+[ tem][per] -> [ temper] 3174
+[emm][ick] -> [emmick] 3175
+[ir][st] -> [irst] 3176
+[pl][es] -> [ples] 3177
+[ disc][uss] -> [ discuss] 3178
+[ cont][ain] -> [ contain] 3179
+[ೋ][ಮ] -> [ೋಮ] 3180
+[ comple][te] -> [ complete] 3181
+[ri][e] -> [rie] 3182
+[o][le] -> [ole] 3183
+[ ac][ross] -> [ across] 3184
+[ com][b] -> [ comb] 3185
+[it][or] -> [itor] 3186
+[ es][pecially] -> [ especially] 3187
+[ commun][ity] -> [ community] 3188
+[ಂದ][್ರ] -> [ಂದ್ರ] 3189
+[ ಅವರ][ನ್ನು] -> [ ಅವರನ್ನು] 3190
+[ಬ್ಲ][್ಯೂ] -> [ಬ್ಲ್ಯೂ] 3191
+[ 200][5] -> [ 2005] 3192
+[ im][age] -> [ image] 3193
+[ द][क्षिण] -> [ दक्षिण] 3194
+[ ][z] -> [ z] 3195
+[ ಪ್ರಮ][ುಖ] -> [ ಪ್ರಮುಖ] 3196
+[ृ][ंख] -> [ृंख] 3197
+[ा�][�] -> [ाख] 3198
+[ ग][ुल] -> [ गुल] 3199
+[ 201][0] -> [ 2010] 3200
+[ e][ither] -> [ either] 3201
+[ be][gan] -> [ began] 3202
+[ N][ational] -> [ National] 3203
+[्�][�ा] -> [्या] 3204
+[ cle][ar] -> [ clear] 3205
+[ E][st] -> [ Est] 3206
+[ m][ult] -> [ mult] 3207
+[ot][e] -> [ote] 3208
+[ys][is] -> [ysis] 3209
+[ फिल][्म] -> [ फिल्म] 3210
+[ आ][ध] -> [ आध] 3211
+[am][ed] -> [amed] 3212
+[av][es] -> [aves] 3213
+[ो][च] -> [ोच] 3214
+[ॉ][स] -> [ॉस] 3215
+[al][m] -> [alm] 3216
+[ d][ark] -> [ dark] 3217
+[ t][akes] -> [ takes] 3218
+[ ಸ್ಥ][ಾಪ] -> [ ಸ್ಥಾಪ] 3219
+[b][y] -> [by] 3220
+[ grou][ps] -> [ groups] 3221
+[ pro][ced] -> [ proced] 3222
+[ ಪ್ರ][ಸ] -> [ ಪ್ರಸ] 3223
+[ುತ್ತ][ವೆ] -> [ುತ್ತವೆ] 3224
+[ದ][ಲು] -> [ದಲು] 3225
+[ 201][3] -> [ 2013] 3226
+[ 201][1] -> [ 2011] 3227
+[ ಮಾರ][್ಗ] -> [ ಮಾರ್ಗ] 3228
+[ದ][್ದ] -> [ದ್ದ] 3229
+[ು][ಂಬ] -> [ುಂಬ] 3230
+[ ಬ][ಿಡ] -> [ ಬಿಡ] 3231
+[ beh][av] -> [ behav] 3232
+[ೆ][ನ್] -> [ೆನ್] 3233
+[ re][d] -> [ red] 3234
+[ श्र][ृंख] -> [ श्रृंख] 3235
+[al][king] -> [alking] 3236
+[ ಸಾಮ][್ರಾಜ] -> [ ಸಾಮ್ರಾಜ] 3237
+[ संग][ीत] -> [ संगीत] 3238
+[ar][l] -> [arl] 3239
+[ ಬಲ್ಗೇರ][ಿಯನ್] -> [ ಬಲ್ಗೇರಿಯನ್] 3240
+[ W][emmick] -> [ Wemmick] 3241
+[s][p] -> [sp] 3242
+[ rel][ated] -> [ related] 3243
+[ as][k] -> [ ask] 3244
+[ su][m] -> [ sum] 3245
+[ p][ath] -> [ path] 3246
+[r][ing] -> [ring] 3247
+[्य][ों] -> [्यों] 3248
+[ ಹ][ಲವ] -> [ ಹಲವ] 3249
+[ ಹಲವ][ಾರು] -> [ ಹಲವಾರು] 3250
+[ ದ][ಕ್ಷ] -> [ ದಕ್ಷ] 3251
+[ ಪ][ಿ] -> [ ಪಿ] 3252
+[it][ing] -> [iting] 3253
+[in][king] -> [inking] 3254
+[ं�][�] -> [ंच] 3255

requirements.txt

@@ -0,0 +1,3 @@
+gradio
+tiktoken
+regex

tokenizer.py

@@ -0,0 +1,91 @@
+import regex as re
+
+def get_stats(ids, freq):
+    for pair in zip(ids[:-1], ids[1:]):
+        freq[pair] = freq.get(pair, 0) + 1
+
+def merge(ids, pair, idx):
+    newids = []
+    i = 0
+    while i < len(ids):
+        if i < len(ids) - 1 and ids[i] == pair[0] and ids[i+1] == pair[1]:
+            newids.append(idx)
+            i += 2
+        else:
+            newids.append(ids[i])
+            i += 1
+    return newids
+
+def _build_vocab(merges):
+    vocab = {idx: bytes([idx]) for idx in range(256)}
+    for (p0, p1), idx in merges.items():
+        vocab[idx] = vocab[p0] + vocab[p1]
+    return vocab
+
+class CustomTokenizer:
+    def __init__(self, model_file):
+        self.merges, self.special_tokens, self.vocab = self.load(model_file)
+        self.pattern = r"""(?i) 's|'t|'re|'ve|'m|'ll|'d| ?\b[\p{L}\u0900-\u0963|\u0966-\u097F]+\b| ?\b[\p{L}\u0C80-\u0C9E|\u0CA0-\u0CFF]+\b| ?[\p{N}]+| ?[.,!?;:'"-]| ?[\u0964-\u0965]| ?[\u0C9E-\u0C9F]| ?[^\s\p{L}\p{N}\u0900-\u097F\u0C80-\u0CFF]+| \s+(?!\S)| \s+"""
+        self.regex = re.compile(self.pattern)
+
+    def load(self, model_file):
+        merges = {}
+        special_tokens = {}
+        idx = 256
+        with open(model_file, 'r', encoding="utf-8") as f:
+            version = f.readline().strip()
+            assert version == "minbpe v1"
+            pattern = f.readline().strip()
+            num_special = int(f.readline().strip())
+            for _ in range(num_special):
+                special, special_idx = f.readline().strip().split()
+                special_tokens[special] = int(special_idx)
+            for line in f:
+                idx1, idx2 = map(int, line.split())
+                merges[(idx1, idx2)] = idx
+                idx += 1
+        vocab = _build_vocab(merges)
+        return merges, special_tokens, vocab
+
+    def _encode_chunk(self, chunk_bytes: bytes) -> list[int]:
+        tokens = list(chunk_bytes)
+        while len(tokens) >= 2:
+            stats = {}
+            get_stats(tokens, stats)
+            pair = min(stats, key=lambda p: self.merges.get(p, float("inf")))
+            if pair not in self.merges:
+                break
+            idx = self.merges[pair]
+            tokens = merge(tokens, pair, idx)
+        return tokens
+
+    def encode(self, text: str, allowed_special="none") -> list[int]:
+        special = {}
+        if allowed_special == "all":
+            special = self.special_tokens
+        elif allowed_special == "none":
+            special = {}
+        elif isinstance(allowed_special, set):
+            special = {k: v for k, v in self.special_tokens.items() if k in allowed_special}
+        
+        if not special:
+            chunk_texts = re.findall(self.regex, text)
+            ids_list = []
+            for text in chunk_texts:
+                chunk_bytes = text.encode("utf-8")
+                ids = self._encode_chunk(chunk_bytes)
+                ids_list.extend(ids)
+            return ids_list
+
+        special_pattern = "(" + "|".join(re.escape(token) for token in special) + ")"
+        parts = re.split(special_pattern, text)
+        ids = []
+        for part in parts:
+            if part in special:
+                ids.append(special[part])
+            else:
+                chunk_texts = re.findall(self.regex, part)
+                for text in chunk_texts:
+                    chunk_bytes = text.encode("utf-8")
+                    ids.extend(self._encode_chunk(chunk_bytes))
+        return ids