change data format to hdfs

Signed-off-by: Diwank Singh <diwank.singh@gmail.com>

.gitattributes

@@ -1,3 +1,4 @@
+*.h5 filter=lfs diff=lfs merge=lfs -text
 *.txt filter=lfs diff=lfs merge=lfs -text
 *.json filter=lfs diff=lfs merge=lfs -text
 *.xml filter=lfs diff=lfs merge=lfs -text

data/crowd_transliteration/{crowd_transliterations.hi-en.txt → data.h5}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2e45d29dec243dea84f0d233e453bc14efe48d795da4258e64aba9144d36365f
-size 380351
+oid sha256:3526e48f077b7dfafbdedcbbed739675cceaacef2a31ab651802afc4404300e8
+size 3634483

data/crowd_transliteration/process.ipynb

@@ -0,0 +1,178 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "File ‘crowd_transliterations.hi-en.txt’ already there; not retrieving.\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -nc https://raw.githubusercontent.com/chsasank/indic-transliteration/master/data/crowd_transliterations.hi-en.txt"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Index(['target_hinglish', 'source_hindi'], dtype='object')"
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "df = pd.read_csv(\"./crowd_transliterations.hi-en.txt\", sep='\\t', names=[\"target_hinglish\", \"source_hindi\"])\n",
+    "df.columns"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {},
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **crowd_transliterations.hi-en.txt**, only `target_hinglish` and `source_hindi` are valid"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "b54fdd52-1ab0-4c84-89e5-0bcb8fcbfbeb",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Add empty columns\n",
+    "df[\"raw_input\"] = \\\n",
+    "  df[\"parallel_english\"] = \\\n",
+    "  df[\"alternates\"] = \\\n",
+    "  df[\"annotations\"] = None\n",
+    "\n",
+    "# Split dataset\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "_train_eval_df, test_df = train_test_split(df, test_size=0.1)\n",
+    "train_df, eval_df = train_test_split(_train_eval_df, test_size=0.1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Requirement already satisfied: tables in /opt/conda/lib/python3.7/site-packages (3.7.0)\n",
+      "Requirement already satisfied: numexpr>=2.6.2 in /opt/conda/lib/python3.7/site-packages (from tables) (2.8.1)\n",
+      "Requirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['target_hinglish', 'source_hindi', 'raw_input', 'parallel_english',\n",
+      "       'alternates', 'annotations'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "3298f2f3-3e21-478e-b027-947c992f880d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm crowd_transliterations.hi-en.txt"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

data/hindi_romanized_dump/{hi_rom.txt → data.h5}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:964952e12f06aee34c9a689636bd56bca6541b78558975f7ee15b277a9d31ca4
-size 517324149
+oid sha256:c67ea7c1fd0400f01f2be2779af5d8d112936cf99ba99c11fa12c73fb95873c5
+size 579369992

data/hindi_romanized_dump/process.ipynb

@@ -0,0 +1,204 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "File ‘hi_rom.txt.xz’ already there; not retrieving.\n",
+      "\n",
+      "hi_rom.txt.xz (1/1)\n",
+      "xz: hi_rom.txt: File exists\n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -nc http://data.statmt.org/cc-100/hi_rom.txt.xz\n",
+    "!xz -d -v hi_rom.txt.xz\n",
+    "!rm hi_rom.txt.xz"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {
+    "tags": []
+   },
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **hi_rom.txt**, only `target_hinglish` is valid"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Requirement already satisfied: clean-text[gpl] in /opt/conda/lib/python3.7/site-packages (0.6.0)\n",
+      "Requirement already satisfied: tqdm in /opt/conda/lib/python3.7/site-packages (4.62.3)\n",
+      "Requirement already satisfied: emoji<2.0.0,>=1.0.0 in /opt/conda/lib/python3.7/site-packages (from clean-text[gpl]) (1.6.3)\n",
+      "Requirement already satisfied: ftfy<7.0,>=6.0 in /opt/conda/lib/python3.7/site-packages (from clean-text[gpl]) (6.1.1)\n",
+      "Requirement already satisfied: unidecode<2.0.0,>=1.1.1 in /opt/conda/lib/python3.7/site-packages (from clean-text[gpl]) (1.3.2)\n",
+      "Requirement already satisfied: wcwidth>=0.2.5 in /opt/conda/lib/python3.7/site-packages (from ftfy<7.0,>=6.0->clean-text[gpl]) (0.2.5)\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "10251114it [20:30, 8333.31it/s] \n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install clean-text[gpl] tqdm\n",
+    "import pandas as pd\n",
+    "from tqdm import tqdm\n",
+    "from cleantext import clean as clean_\n",
+    "\n",
+    "clean = lambda x: clean_(\n",
+    "    x, \n",
+    "    no_line_breaks=True, no_punct=True,\n",
+    "    no_urls=True, replace_with_url=\"<|url|>\",\n",
+    "    no_emoji=True, no_phone_numbers=True, replace_with_phone_number=\"<|phonenumber|>\",\n",
+    "    no_emails=True, replace_with_email=\"<|email|>\",\n",
+    "    no_currency_symbols=True, replace_with_currency_symbol=\"\", )\n",
+    "\n",
+    "with open(\"./hi_rom.txt\", 'r') as file:\n",
+    "    df = pd.DataFrame(\n",
+    "        [(clean(line), None, None, None, None, None) for line in tqdm(file)],\n",
+    "        columns=[\"target_hinglish\", \"source_hindi\", \"parallel_english\", \"annotations\", \"raw_input\", \"alternates\"] )"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "b54fdd52-1ab0-4c84-89e5-0bcb8fcbfbeb",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Split dataset\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "_train_eval_df, test_df = train_test_split(df, test_size=0.1)\n",
+    "train_df, eval_df = train_test_split(_train_eval_df, test_size=0.1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting tables\n",
+      "  Using cached tables-3.7.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
+      "Requirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Collecting numexpr>=2.6.2\n",
+      "  Using cached numexpr-2.8.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (379 kB)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n",
+      "Installing collected packages: numexpr, tables\n",
+      "Successfully installed numexpr-2.8.1 tables-3.7.0\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['target_hinglish', 'source_hindi', 'parallel_english', 'annotations',\n",
+      "       'raw_input', 'alternates'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "3298f2f3-3e21-478e-b027-947c992f880d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm hi_rom.txt"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

data/hindi_xlit/HiEn_ann1_train.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1c92b3c8e955dcc3574223b4fc0516dfb3df1205131b3ad0fef00e6bd35745b4
-size 2993734

data/hindi_xlit/HiEn_ann1_valid.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4c9336adbabcfcd8602eee9866f8193e4d38300d933217fe3eb77ffd6b993f4c
-size 335141

data/hindi_xlit/data.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c79a4fab51fc5303da465270be9994a13aa803d04eb7c452ac488239d47eadcf
+size 10892954

data/hindi_xlit/process.ipynb

@@ -0,0 +1,200 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "--2022-03-05 12:25:22--  https://github.com/AI4Bharat/IndianNLP-Transliteration/releases/download/DATA/Hindi_Xlit_dataset.zip\n",
+      "Resolving github.com (github.com)... 140.82.114.3\n",
+      "Connecting to github.com (github.com)|140.82.114.3|:443... connected.\n",
+      "HTTP request sent, awaiting response... 302 Found\n",
+      "Location: https://objects.githubusercontent.com/github-production-release-asset-2e65be/231321785/14c95280-01a2-11eb-921f-4221081fa4b2?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIAIWNJYAX4CSVEH53A%2F20220305%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20220305T122522Z&X-Amz-Expires=300&X-Amz-Signature=ef0c94bb0f3602f5edbca49df20bb64a477fe34bfec15b5ee78b43f9be4da4e6&X-Amz-SignedHeaders=host&actor_id=0&key_id=0&repo_id=231321785&response-content-disposition=attachment%3B%20filename%3DHindi_Xlit_dataset.zip&response-content-type=application%2Foctet-stream [following]\n",
+      "--2022-03-05 12:25:22--  https://objects.githubusercontent.com/github-production-release-asset-2e65be/231321785/14c95280-01a2-11eb-921f-4221081fa4b2?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIAIWNJYAX4CSVEH53A%2F20220305%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20220305T122522Z&X-Amz-Expires=300&X-Amz-Signature=ef0c94bb0f3602f5edbca49df20bb64a477fe34bfec15b5ee78b43f9be4da4e6&X-Amz-SignedHeaders=host&actor_id=0&key_id=0&repo_id=231321785&response-content-disposition=attachment%3B%20filename%3DHindi_Xlit_dataset.zip&response-content-type=application%2Foctet-stream\n",
+      "Resolving objects.githubusercontent.com (objects.githubusercontent.com)... 185.199.108.133, 185.199.110.133, 185.199.111.133, ...\n",
+      "Connecting to objects.githubusercontent.com (objects.githubusercontent.com)|185.199.108.133|:443... connected.\n",
+      "HTTP request sent, awaiting response... 200 OK\n",
+      "Length: 609266 (595K) [application/octet-stream]\n",
+      "Saving to: ‘Hindi_Xlit_dataset.zip’\n",
+      "\n",
+      "Hindi_Xlit_dataset. 100%[===================>] 594.99K  --.-KB/s    in 0.04s   \n",
+      "\n",
+      "2022-03-05 12:25:23 (13.8 MB/s) - ‘Hindi_Xlit_dataset.zip’ saved [609266/609266]\n",
+      "\n",
+      "Archive:  ./Hindi_Xlit_dataset.zip\n",
+      "  inflating: HiEn_ann1_test.json     \n",
+      "  inflating: HiEn_ann1_train.json    \n",
+      "  inflating: HiEn_ann1_valid.json    \n",
+      "  inflating: legalcode.txt           \n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -nc https://github.com/AI4Bharat/IndianNLP-Transliteration/releases/download/DATA/Hindi_Xlit_dataset.zip\n",
+    "!unzip -n ./Hindi_Xlit_dataset.zip\n",
+    "!rm ./legalcode.txt ./Hindi_Xlit_dataset.zip"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {},
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **HiEn_ann1**, only `target_hinglish` and `source_hindi` are valid"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import json\n",
+    "\n",
+    "with open(\"./HiEn_ann1_train.json\", 'r') as f:\n",
+    "    train_data = json.load(f)\n",
+    "\n",
+    "with open(\"./HiEn_ann1_valid.json\", 'r') as f:\n",
+    "    eval_data = json.load(f)\n",
+    "\n",
+    "with open(\"./HiEn_ann1_test.json\", 'r') as f:\n",
+    "    test_data = json.load(f)\n",
+    "\n",
+    "train_df = pd.DataFrame(\n",
+    "    [(source_hindi, target_hinglish, None, None, None, None) \n",
+    "     for source_hindi, values in train_data.items() \n",
+    "     for target_hinglish in values ], \n",
+    "    columns=[\"source_hindi\", \"target_hinglish\", \"parallel_english\", \"annotations\", \"raw_input\", \"alternates\"] )\n",
+    "\n",
+    "eval_df = pd.DataFrame(\n",
+    "    [(source_hindi, target_hinglish, None, None, None, None) \n",
+    "     for source_hindi, values in eval_data.items() \n",
+    "     for target_hinglish in values ], \n",
+    "    columns=[\"source_hindi\", \"target_hinglish\", \"parallel_english\", \"annotations\", \"raw_input\", \"alternates\"] )\n",
+    "\n",
+    "test_df = pd.DataFrame(\n",
+    "    [(source_hindi, target_hinglish, None, None, None, None) \n",
+    "     for source_hindi, values in test_data.items() \n",
+    "     for target_hinglish in values ], \n",
+    "    columns=[\"source_hindi\", \"target_hinglish\", \"parallel_english\", \"annotations\", \"raw_input\", \"alternates\"] )\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting tables\n",
+      "  Using cached tables-3.7.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
+      "Requirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Collecting numexpr>=2.6.2\n",
+      "  Using cached numexpr-2.8.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (379 kB)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n",
+      "Installing collected packages: numexpr, tables\n",
+      "Successfully installed numexpr-2.8.1 tables-3.7.0\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['source_hindi', 'target_hinglish', 'parallel_english', 'annotations',\n",
+      "       'raw_input', 'alternates'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "3298f2f3-3e21-478e-b027-947c992f880d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm HiEn_ann1_test.json\n",
+    "!rm HiEn_ann1_train.json\n",
+    "!rm HiEn_ann1_valid.json"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

data/hinge/.ipynb_checkpoints/eval_synthetic-checkpoint.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:01179fcb3c0dfbf45cddd0c89cd40a867676668d10b5cdea29f6f2a765d2ec82
-size 67597

data/{fire2013/HindiEnglish_FIRE2013_AnnotatedDev.txt → hinge/data.h5}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6251057ea0c2e2e1f2daa9d96a0726f4bb713762296b3eb96ed6dfbab0b5dba6
-size 54391
+oid sha256:f4a0a15fa9c69d5a705707c73d1fea20abf9882fad18219579e4660a0fd213ea
+size 5243006

data/hinge/eval_human.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a48b0adeaa5a65f097ff644f695b20ef6227ef7ff46bd0c573b84186130ff49c
-size 139425

data/hinge/eval_human.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:a48b0adeaa5a65f097ff644f695b20ef6227ef7ff46bd0c573b84186130ff49c
-size 139425

data/hinge/eval_synthetic.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:6d6c862f0d2654b08a20a60cd4d4d68b041da9c3c0029fd91e52e9fd3cabe240
-size 176109

data/hinge/process.ipynb

@@ -0,0 +1,169 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Please request data from https://codalab.lisn.upsaclay.fr/competitions/1688\n",
+      "We only need eval_synthetic.csv and train_synthetic.csv\n",
+      "eval_synthetic.csv  process.ipynb  train_synthetic.csv\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(\"Please request data from https://codalab.lisn.upsaclay.fr/competitions/1688\")\n",
+    "print(\"We only need eval_synthetic.csv and train_synthetic.csv\")\n",
+    "!ls ."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {},
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **HingE**, only `target_hinglish`, `parallel_english` and `source_hindi` are valid"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "train_df = pd.read_csv(\"./train_synthetic.csv\", names=[\"parallel_english\", \"source_hindi\", \"target_hinglish\"], header=0, usecols=[0, 1, 2])\n",
+    "_test_eval_df = pd.read_csv(\"./train_synthetic.csv\", names=[\"parallel_english\", \"source_hindi\", \"target_hinglish\"], header=0, usecols=[0, 1, 2])\n",
+    "\n",
+    "# Add empty columns\n",
+    "train_df[\"raw_input\"] = \\\n",
+    "  train_df[\"alternates\"] = \\\n",
+    "  train_df[\"annotations\"] = None\n",
+    "\n",
+    "_test_eval_df[\"raw_input\"] = \\\n",
+    "  _test_eval_df[\"alternates\"] = \\\n",
+    "  _test_eval_df[\"annotations\"] = None\n",
+    "\n",
+    "# Split dataset\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "eval_df, test_df = train_test_split(_test_eval_df, test_size=0.5)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting tables\n",
+      "  Using cached tables-3.7.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
+      "Collecting numexpr>=2.6.2\n",
+      "  Using cached numexpr-2.8.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (379 kB)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Requirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n",
+      "Installing collected packages: numexpr, tables\n",
+      "Successfully installed numexpr-2.8.1 tables-3.7.0\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['parallel_english', 'source_hindi', 'target_hinglish', 'raw_input',\n",
+      "       'alternates', 'annotations'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "3298f2f3-3e21-478e-b027-947c992f880d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm eval_synthetic.csv\n",
+    "!rm train_synthetic.csv"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

data/hinge/train_human.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8fe6085bbcb39c014f79faaac540ce7fa03d054924a600f340247f9ebb0dcf21
-size 675953

data/hinge/train_human.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8fe6085bbcb39c014f79faaac540ce7fa03d054924a600f340247f9ebb0dcf21
-size 675953

data/hinge/train_synthetic.csv

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1e7b09167b56515a617a90f8b83c61711df9e2e3796557f7c54dc4f7e036a666
-size 1265593

data/{fire2013/HindiEnglish_FIRE2013_Test_GT.txt → hinglish_norm/data.h5}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c31a05d3884a680f9507b7101d1c7d33fb62d870e568ce25cb5b763574c90aae
-size 29326
+oid sha256:ebc7cf76907c37dc6f726b69a6c8532628680fcba5ae1cb13609707180beb3ec
+size 4837335

data/hinglish_norm/hinglishNorm_trainSet.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cf622e18d943de8fc28a869b0a7d76076676b83238ebbd61f96850ef86c2f337
-size 2370544

data/hinglish_norm/process.ipynb

@@ -0,0 +1,201 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "File ‘hinglishNorm.json’ already there; not retrieving.\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "!wget -nc https://raw.githubusercontent.com/piyushmakhija5/hinglishNorm/master/dataset/hinglishNorm.json"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Index(['id', 'inputText', 'tags', 'normalizedText'], dtype='object')"
+      ]
+     },
+     "execution_count": 2,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "df = pd.read_json(\"./hinglishNorm.json\")\n",
+    "df.columns"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {},
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **hinglishNorm**, only `target_hinglish`, `raw_input` and `annotations` are valid\n",
+    "\n",
+    "### Mappings\n",
+    "- `normalizedText` _=>_ `target_hinglish`\n",
+    "- `inputText` _=>_ `raw_input`\n",
+    "- `tags` _=>_ `annotations` (after json.loads)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "b54fdd52-1ab0-4c84-89e5-0bcb8fcbfbeb",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Add empty columns\n",
+    "df[\"source_hindi\"] = \\\n",
+    "  df[\"parallel_english\"] = \\\n",
+    "  df[\"alternates\"] = None\n",
+    "\n",
+    "# Remove unnecessary columns\n",
+    "df = df.drop(\"id\", axis=1)\n",
+    "\n",
+    "# Rename columns\n",
+    "df = df.rename(columns={\n",
+    "    \"normalizedText\": \"target_hinglish\", \n",
+    "    \"inputText\": \"raw_input\", \n",
+    "    \"tags\": \"annotations\", })\n",
+    "\n",
+    "# Parse annotations json\n",
+    "import json\n",
+    "df[\"annotations\"] = df[\"annotations\"].map(lambda x: json.loads(x.replace(\"'\", '\"')))\n",
+    "\n",
+    "# Split dataset\n",
+    "from sklearn.model_selection import train_test_split\n",
+    "_train_eval_df, test_df = train_test_split(df, test_size=0.1)\n",
+    "train_df, eval_df = train_test_split(_train_eval_df, test_size=0.1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting tables\n",
+      "  Downloading tables-3.7.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
+      "     |████████████████████████████████| 5.9 MB 4.9 MB/s            \n",
+      "\u001b[?25hRequirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Collecting numexpr>=2.6.2\n",
+      "  Downloading numexpr-2.8.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (379 kB)\n",
+      "     |████████████████████████████████| 379 kB 73.4 MB/s            \n",
+      "\u001b[?25hRequirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n",
+      "Installing collected packages: numexpr, tables\n",
+      "Successfully installed numexpr-2.8.1 tables-3.7.0\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['raw_input', 'annotations', 'target_hinglish', 'source_hindi',\n",
+      "       'parallel_english', 'alternates'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "c8908455-c76f-4ee2-9608-b215f6fafa7c",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm hinglishNorm.json"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

data/news2018/NEWS2018_M-EnHi_dev.xml

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:754d4174f291a0f6209f6ddccca4a7af04700f6bfd4d9b83ffb24c0711d4aaa3
-size 127866

data/news2018/NEWS2018_M-EnHi_trn.xml

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:32a0f41ebe7baa635d0fc2ada5feb1ad93e18a03a22dea8f7bc015cab5adc0b6
-size 1681958

data/news2018/NEWS2018_M-EnHi_tst.xml

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:57db5526c30ec34cd707ed365558231283788b34be4a77ae2b807f2867e6be30
-size 60532

data/{hindi_xlit/HiEn_ann1_test.json → news2018/data.h5}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8e8f8b749849ed6c69220b46da33a27f7a2f7e97e053935b834c5e2ae68600e7
-size 338964
+oid sha256:692631290d13d664e4ea55f1fbfb6d6138dd7836c08dea658fabf36ed70beda1
+size 3820798

data/news2018/process.ipynb

@@ -0,0 +1,188 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "bed45d12-7681-4ba4-9c89-48a3515704e2",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Please request data from http://workshop.colips.org/news2018/terms&conditions_msri.html\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(\"Please request data from http://workshop.colips.org/news2018/terms&conditions_msri.html\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "b5c7c7c7-b9a6-4ea2-a5ef-edaf982ae0ad",
+   "metadata": {},
+   "source": [
+    "### Required columns\n",
+    "- target_hinglish\n",
+    "- source_hindi\n",
+    "- parallel_english\n",
+    "- annotations\n",
+    "- raw_input\n",
+    "- alternates\n",
+    "\n",
+    "> For **NEWS2018**, only `target_hinglish` and `source_hindi` are valid"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "965589a9-c62e-4659-a6bc-6f0a2bad5d19",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Requirement already satisfied: lxml in /opt/conda/lib/python3.7/site-packages (4.8.0)\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install lxml\n",
+    "import pandas as pd\n",
+    "\n",
+    "train_df = pd.read_xml(\"./NEWS2018_M-EnHi_trn.xml\", names=[\"target_hinglish\", \"source_hindi\"], elems_only=True)\n",
+    "test_df = pd.read_xml(\"./NEWS2018_M-EnHi_tst.xml\", names=[\"target_hinglish\", \"source_hindi\"], elems_only=True)\n",
+    "eval_df = pd.read_xml(\"./NEWS2018_M-EnHi_dev.xml\", names=[\"target_hinglish\", \"source_hindi\"], elems_only=True)\n",
+    "\n",
+    "# Add empty columns\n",
+    "train_df[\"raw_input\"] = \\\n",
+    "  train_df[\"parallel_english\"] = \\\n",
+    "  train_df[\"alternates\"] = \\\n",
+    "  train_df[\"annotations\"] = None\n",
+    "\n",
+    "# Add empty columns\n",
+    "test_df[\"raw_input\"] = \\\n",
+    "  test_df[\"parallel_english\"] = \\\n",
+    "  test_df[\"alternates\"] = \\\n",
+    "  test_df[\"annotations\"] = None\n",
+    "\n",
+    "# Add empty columns\n",
+    "eval_df[\"raw_input\"] = \\\n",
+    "  eval_df[\"parallel_english\"] = \\\n",
+    "  eval_df[\"alternates\"] = \\\n",
+    "  eval_df[\"annotations\"] = None\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "6e804366-34cd-45c7-b3c6-46b7b8c1b420",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting tables\n",
+      "  Using cached tables-3.7.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (5.9 MB)\n",
+      "Requirement already satisfied: numpy>=1.19.0 in /opt/conda/lib/python3.7/site-packages (from tables) (1.19.5)\n",
+      "Collecting numexpr>=2.6.2\n",
+      "  Using cached numexpr-2.8.1-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (379 kB)\n",
+      "Requirement already satisfied: packaging in /opt/conda/lib/python3.7/site-packages (from tables) (21.3)\n",
+      "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/lib/python3.7/site-packages (from packaging->tables) (3.0.6)\n",
+      "Installing collected packages: numexpr, tables\n",
+      "Successfully installed numexpr-2.8.1 tables-3.7.0\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['target_hinglish', 'source_hindi', 'raw_input', 'parallel_english',\n",
+      "       'alternates', 'annotations'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n",
+      "/opt/conda/lib/python3.7/site-packages/pandas/core/generic.py:2718: PerformanceWarning: \n",
+      "your performance may suffer as PyTables will pickle object types that it cannot\n",
+      "map directly to c-types [inferred_type->mixed,key->block0_values] [items->Index(['target_hinglish', 'raw_input', 'parallel_english', 'alternates',\n",
+      "       'annotations'],\n",
+      "      dtype='object')]\n",
+      "\n",
+      "  encoding=encoding,\n"
+     ]
+    }
+   ],
+   "source": [
+    "!pip install tables\n",
+    "\n",
+    "# Save to hdfs files\n",
+    "train_df.to_hdf(\"./data.h5\", \"train\", complevel=9)\n",
+    "test_df.to_hdf(\"./data.h5\", \"test\", complevel=9)\n",
+    "eval_df.to_hdf(\"./data.h5\", \"eval\", complevel=9)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "3298f2f3-3e21-478e-b027-947c992f880d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Confirm that everything worked as expected\n",
+    "\n",
+    "# Load from hdfs files\n",
+    "_train_df = pd.read_hdf(\"./data.h5\", \"train\")\n",
+    "_test_df = pd.read_hdf(\"./data.h5\", \"test\")\n",
+    "_eval_df = pd.read_hdf(\"./data.h5\", \"eval\")\n",
+    "\n",
+    "assert (len(_train_df) == len(train_df)) == \\\n",
+    "  (len(_eval_df) == len(eval_df)) == \\\n",
+    "  (len(_test_df) == len(test_df))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "60461121-bed5-4ba0-ba7d-dd46256c62e3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!rm NEWS2018_M-EnHi_trn.xml\n",
+    "!rm NEWS2018_M-EnHi_tst.xml\n",
+    "!rm NEWS2018_M-EnHi_dev.xml"
+   ]
+  }
+ ],
+ "metadata": {
+  "environment": {
+   "kernel": "python3",
+   "name": "managed-notebooks.m87",
+   "type": "gcloud",
+   "uri": "gcr.io/deeplearning-platform-release/base-cu110:latest"
+  },
+  "kernelspec": {
+   "display_name": "Python (Local)",
+   "language": "python",
+   "name": "local-base"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.12"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

hinglish-dump.py

@@ -37,9 +37,7 @@ _URLS = {
     ])),
     "hinge": list(map(lambda x: f"{_HOMEPAGE}/resolve/main/data/{x}" , [
         "hinge/eval_human.csv",
-        "hinge/eval_human.pkl",
         "hinge/train_human.csv",
-        "hinge/train_human.pkl",
         "hinge/train_synthetic.csv",
         "hinge/eval_synthetic.csv",
     ])),
@@ -56,7 +54,7 @@ _URLS = {
 config_names = _URLS.keys()
 version = datasets.Version("1.0.0")
 
-class HinglishDumpDataset(datasets.GeneratorBasedBuilder):
+class HinglishDumpDataset(datasets.DatasetBuilder):
     """Raw merged dump of Hinglish (hi-EN) datasets."""
 
     VERSION = version
@@ -66,6 +64,8 @@ class HinglishDumpDataset(datasets.GeneratorBasedBuilder):
         datasets.BuilderConfig(name=subset, version=version, description=f"Config for {subset}")
         for subset in config_names
     ]
+    
+    DEFAULT_CONFIG_NAME = None
 
     def _info(self):
                 
@@ -81,7 +81,8 @@ class HinglishDumpDataset(datasets.GeneratorBasedBuilder):
         # It can accept any type or nested list/dict and will give back the same structure with the url replaced with path to local files.
         # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
 
-        data_dir = dl_manager.download_and_extract(_URLS)
+        data_dir = self.data_dir = dl_manager.download_and_extract(_URLS)
+        
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,