Spaces:
Runtime error
Runtime error
File size: 11,759 Bytes
ed4d993 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 |
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"id": "fc935871-7640-41c6-b798-58514d860fe0",
"metadata": {},
"source": [
"## LLaMA2 chat with SQL\n",
"\n",
"Open source, local LLMs are great to consider for any application that demands data privacy.\n",
"\n",
"SQL is one good example. \n",
"\n",
"This cookbook shows how to perform text-to-SQL using various local versions of LLaMA2 run locally.\n",
"\n",
"## Packages"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "81adcf8b-395a-4f02-8749-ac976942b446",
"metadata": {},
"outputs": [],
"source": [
"! pip install langchain replicate"
]
},
{
"cell_type": "markdown",
"id": "8e13ed66-300b-4a23-b8ac-44df68ee4733",
"metadata": {},
"source": [
"## LLM\n",
"\n",
"There are a few ways to access LLaMA2.\n",
"\n",
"To run locally, we use Ollama.ai. \n",
"\n",
"See [here](/docs/integrations/chat/ollama) for details on installation and setup.\n",
"\n",
"Also, see [here](/docs/guides/development/local_llms) for our full guide on local LLMs.\n",
" \n",
"To use an external API, which is not private, we can use Replicate."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "6a75a5c6-34ee-4ab9-a664-d9b432d812ee",
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Init param `input` is deprecated, please use `model_kwargs` instead.\n"
]
}
],
"source": [
"# Local\n",
"from langchain_community.chat_models import ChatOllama\n",
"\n",
"llama2_chat = ChatOllama(model=\"llama2:13b-chat\")\n",
"llama2_code = ChatOllama(model=\"codellama:7b-instruct\")\n",
"\n",
"# API\n",
"from langchain_community.llms import Replicate\n",
"\n",
"# REPLICATE_API_TOKEN = getpass()\n",
"# os.environ[\"REPLICATE_API_TOKEN\"] = REPLICATE_API_TOKEN\n",
"replicate_id = \"meta/llama-2-13b-chat:f4e2de70d66816a838a89eeeb621910adffb0dd0baba3976c96980970978018d\"\n",
"llama2_chat_replicate = Replicate(\n",
" model=replicate_id, input={\"temperature\": 0.01, \"max_length\": 500, \"top_p\": 1}\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "ce96f7ea-b3d5-44e1-9fa5-a79e04a9e1fb",
"metadata": {},
"outputs": [],
"source": [
"# Simply set the LLM we want to use\n",
"llm = llama2_chat"
]
},
{
"cell_type": "markdown",
"id": "80222165-f353-4e35-a123-5f70fd70c6c8",
"metadata": {},
"source": [
"## DB\n",
"\n",
"Connect to a SQLite DB.\n",
"\n",
"To create this particular DB, you can use the code and follow the steps shown [here](https://github.com/facebookresearch/llama-recipes/blob/main/demo_apps/StructuredLlama.ipynb)."
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "025bdd82-3bb1-4948-bc7c-c3ccd94fd05c",
"metadata": {},
"outputs": [],
"source": [
"from langchain_community.utilities import SQLDatabase\n",
"\n",
"db = SQLDatabase.from_uri(\"sqlite:///nba_roster.db\", sample_rows_in_table_info=0)\n",
"\n",
"\n",
"def get_schema(_):\n",
" return db.get_table_info()\n",
"\n",
"\n",
"def run_query(query):\n",
" return db.run(query)"
]
},
{
"cell_type": "markdown",
"id": "654b3577-baa2-4e12-a393-f40e5db49ac7",
"metadata": {},
"source": [
"## Query a SQL Database \n",
"\n",
"Follow the runnables workflow [here](https://python.langchain.com/docs/expression_language/cookbook/sql_db)."
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "5a4933ea-d9c0-4b0a-8177-ba4490c6532b",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"' SELECT \"Team\" FROM nba_roster WHERE \"NAME\" = \\'Klay Thompson\\';'"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Prompt\n",
"from langchain_core.prompts import ChatPromptTemplate\n",
"\n",
"# Update the template based on the type of SQL Database like MySQL, Microsoft SQL Server and so on\n",
"template = \"\"\"Based on the table schema below, write a SQL query that would answer the user's question:\n",
"{schema}\n",
"\n",
"Question: {question}\n",
"SQL Query:\"\"\"\n",
"prompt = ChatPromptTemplate.from_messages(\n",
" [\n",
" (\"system\", \"Given an input question, convert it to a SQL query. No pre-amble.\"),\n",
" (\"human\", template),\n",
" ]\n",
")\n",
"\n",
"# Chain to query\n",
"from langchain_core.output_parsers import StrOutputParser\n",
"from langchain_core.runnables import RunnablePassthrough\n",
"\n",
"sql_response = (\n",
" RunnablePassthrough.assign(schema=get_schema)\n",
" | prompt\n",
" | llm.bind(stop=[\"\\nSQLResult:\"])\n",
" | StrOutputParser()\n",
")\n",
"\n",
"sql_response.invoke({\"question\": \"What team is Klay Thompson on?\"})"
]
},
{
"cell_type": "markdown",
"id": "a0e9e2c8-9b88-4853-ac86-001bc6cc6695",
"metadata": {},
"source": [
"We can review the results:\n",
"\n",
"* [LangSmith trace](https://smith.langchain.com/public/afa56a06-b4e2-469a-a60f-c1746e75e42b/r) LLaMA2-13 Replicate API\n",
"* [LangSmith trace](https://smith.langchain.com/public/2d4ecc72-6b8f-4523-8f0b-ea95c6b54a1d/r) LLaMA2-13 local \n"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "2a2825e3-c1b6-4f7d-b9c9-d9835de323bb",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"AIMessage(content=' Based on the table schema and SQL query, there are 30 unique teams in the NBA.')"
]
},
"execution_count": 15,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Chain to answer\n",
"template = \"\"\"Based on the table schema below, question, sql query, and sql response, write a natural language response:\n",
"{schema}\n",
"\n",
"Question: {question}\n",
"SQL Query: {query}\n",
"SQL Response: {response}\"\"\"\n",
"prompt_response = ChatPromptTemplate.from_messages(\n",
" [\n",
" (\n",
" \"system\",\n",
" \"Given an input question and SQL response, convert it to a natural language answer. No pre-amble.\",\n",
" ),\n",
" (\"human\", template),\n",
" ]\n",
")\n",
"\n",
"full_chain = (\n",
" RunnablePassthrough.assign(query=sql_response)\n",
" | RunnablePassthrough.assign(\n",
" schema=get_schema,\n",
" response=lambda x: db.run(x[\"query\"]),\n",
" )\n",
" | prompt_response\n",
" | llm\n",
")\n",
"\n",
"full_chain.invoke({\"question\": \"How many unique teams are there?\"})"
]
},
{
"cell_type": "markdown",
"id": "ec17b3ee-6618-4681-b6df-089bbb5ffcd7",
"metadata": {},
"source": [
"We can review the results:\n",
"\n",
"* [LangSmith trace](https://smith.langchain.com/public/10420721-746a-4806-8ecf-d6dc6399d739/r) LLaMA2-13 Replicate API\n",
"* [LangSmith trace](https://smith.langchain.com/public/5265ebab-0a22-4f37-936b-3300f2dfa1c1/r) LLaMA2-13 local "
]
},
{
"cell_type": "markdown",
"id": "1e85381b-1edc-4bb3-a7bd-2ab23f81e54d",
"metadata": {},
"source": [
"## Chat with a SQL DB \n",
"\n",
"Next, we can add memory."
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "022868f2-128e-42f5-8d90-d3bb2f11d994",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"' SELECT \"Team\" FROM nba_roster WHERE \"NAME\" = \\'Klay Thompson\\';'"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Prompt\n",
"from langchain.memory import ConversationBufferMemory\n",
"from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder\n",
"\n",
"template = \"\"\"Given an input question, convert it to a SQL query. No pre-amble. Based on the table schema below, write a SQL query that would answer the user's question:\n",
"{schema}\n",
"\"\"\"\n",
"prompt = ChatPromptTemplate.from_messages(\n",
" [\n",
" (\"system\", template),\n",
" MessagesPlaceholder(variable_name=\"history\"),\n",
" (\"human\", \"{question}\"),\n",
" ]\n",
")\n",
"\n",
"memory = ConversationBufferMemory(return_messages=True)\n",
"\n",
"# Chain to query with memory\n",
"from langchain_core.runnables import RunnableLambda\n",
"\n",
"sql_chain = (\n",
" RunnablePassthrough.assign(\n",
" schema=get_schema,\n",
" history=RunnableLambda(lambda x: memory.load_memory_variables(x)[\"history\"]),\n",
" )\n",
" | prompt\n",
" | llm.bind(stop=[\"\\nSQLResult:\"])\n",
" | StrOutputParser()\n",
")\n",
"\n",
"\n",
"def save(input_output):\n",
" output = {\"output\": input_output.pop(\"output\")}\n",
" memory.save_context(input_output, output)\n",
" return output[\"output\"]\n",
"\n",
"\n",
"sql_response_memory = RunnablePassthrough.assign(output=sql_chain) | save\n",
"sql_response_memory.invoke({\"question\": \"What team is Klay Thompson on?\"})"
]
},
{
"cell_type": "code",
"execution_count": 21,
"id": "800a7a3b-f411-478b-af51-2310cd6e0425",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"AIMessage(content=' Sure! Here\\'s the natural language response based on the given input:\\n\\n\"Klay Thompson\\'s salary is $43,219,440.\"')"
]
},
"execution_count": 21,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Chain to answer\n",
"template = \"\"\"Based on the table schema below, question, sql query, and sql response, write a natural language response:\n",
"{schema}\n",
"\n",
"Question: {question}\n",
"SQL Query: {query}\n",
"SQL Response: {response}\"\"\"\n",
"prompt_response = ChatPromptTemplate.from_messages(\n",
" [\n",
" (\n",
" \"system\",\n",
" \"Given an input question and SQL response, convert it to a natural language answer. No pre-amble.\",\n",
" ),\n",
" (\"human\", template),\n",
" ]\n",
")\n",
"\n",
"full_chain = (\n",
" RunnablePassthrough.assign(query=sql_response_memory)\n",
" | RunnablePassthrough.assign(\n",
" schema=get_schema,\n",
" response=lambda x: db.run(x[\"query\"]),\n",
" )\n",
" | prompt_response\n",
" | llm\n",
")\n",
"\n",
"full_chain.invoke({\"question\": \"What is his salary?\"})"
]
},
{
"cell_type": "markdown",
"id": "b77fee61-f4da-4bb1-8285-14101e505518",
"metadata": {},
"source": [
"Here is the [trace](https://smith.langchain.com/public/54794d18-2337-4ce2-8b9f-3d8a2df89e51/r)."
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"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": "3.9.16"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|