more complex sql chain (#619)

add a more complex sql chain that first subsets the necessary tables

docs/modules/chains/examples/sqlite.ipynb

@@ -179,10 +179,80 @@
     "db_chain.run(\"How many employees are there in the foobar table?\")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "c12ae15a",
+   "metadata": {},
+   "source": [
+    "## SQLDatabaseSequentialChain\n",
+    "\n",
+    "Chain for querying SQL database that is a sequential chain.\n",
+    "\n",
+    "The chain is as follows:\n",
+    "\n",
+    "    1. Based on the query, determine which tables to use.\n",
+    "    2. Based on those tables, call the normal SQL database chain.\n",
+    "\n",
+    "This is useful in cases where the number of tables in the database is large."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "e59a4740",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from langchain.chains.sql_database.base import SQLDatabaseSequentialChain"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "58bb49b6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "chain = SQLDatabaseSequentialChain.from_llm(llm, db, verbose=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "95017b1a",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "\n",
+      "\u001b[1m> Entering new SQLDatabaseSequentialChain chain...\u001b[0m\n",
+      "Table names to use:\n",
+      "\u001b[33;1m\u001b[1;3m['Employee', 'Customer']\u001b[0m\n",
+      "\u001b[1m> Finished chain.\u001b[0m\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "' 0 employees are also customers.'"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "chain.run(\"How many employees are also customers?\")"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "e59a4740",
+   "id": "b2998b03",
    "metadata": {},
    "outputs": [],
    "source": []