mirror of https://github.com/hwchase17/langchain
Add LCEL to output parser doc (#11880)
Bagatur
9 months ago
committed by
GitHub
parent
049a0357e7
commit
e3664272f0
@ -0,0 +1,259 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "raw",
|
||||
"id": "38831021-76ed-48b3-9f62-d1241a68b6ad",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"---\n",
|
||||
"sidebar_position: 2\n",
|
||||
"title: Output parsers\n",
|
||||
"---"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "a745f98b-c495-44f6-a882-757c38992d76",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"Language models output text. But many times you may want to get more structured information than just text back. This is where output parsers come in.\n",
|
||||
"\n",
|
||||
"Output parsers are classes that help structure language model responses. There are two main methods an output parser must implement:\n",
|
||||
"\n",
|
||||
"- \"Get format instructions\": A method which returns a string containing instructions for how the output of a language model should be formatted.\n",
|
||||
"- \"Parse\": A method which takes in a string (assumed to be the response from a language model) and parses it into some structure.\n",
|
||||
"\n",
|
||||
"And then one optional one:\n",
|
||||
"\n",
|
||||
"- \"Parse with prompt\": A method which takes in a string (assumed to be the response from a language model) and a prompt (assumed to be the prompt that generated such a response) and parses it into some structure. The prompt is largely provided in the event the OutputParser wants to retry or fix the output in some way, and needs information from the prompt to do so.\n",
|
||||
"\n",
|
||||
"## Get started\n",
|
||||
"\n",
|
||||
"Below we go over the main type of output parser, the `PydanticOutputParser`."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 6,
|
||||
"id": "1594b2bf-2a6f-47bb-9a81-38930f8e606b",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"Joke(setup='Why did the chicken cross the road?', punchline='To get to the other side!')"
|
||||
]
|
||||
},
|
||||
"execution_count": 6,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"from typing import List\n",
|
||||
"\n",
|
||||
"from langchain.llms import OpenAI\n",
|
||||
"from langchain.output_parsers import PydanticOutputParser\n",
|
||||
"from langchain.prompts import PromptTemplate\n",
|
||||
"from langchain.pydantic_v1 import BaseModel, Field, validator\n",
|
||||
"\n",
|
||||
"\n",
|
||||
"model = OpenAI(model_name='text-davinci-003', temperature=0.0)\n",
|
||||
"\n",
|
||||
"# Define your desired data structure.\n",
|
||||
"class Joke(BaseModel):\n",
|
||||
" setup: str = Field(description=\"question to set up a joke\")\n",
|
||||
" punchline: str = Field(description=\"answer to resolve the joke\")\n",
|
||||
"\n",
|
||||
" # You can add custom validation logic easily with Pydantic.\n",
|
||||
" @validator('setup')\n",
|
||||
" def question_ends_with_question_mark(cls, field):\n",
|
||||
" if field[-1] != '?':\n",
|
||||
" raise ValueError(\"Badly formed question!\")\n",
|
||||
" return field\n",
|
||||
"\n",
|
||||
"# Set up a parser + inject instructions into the prompt template.\n",
|
||||
"parser = PydanticOutputParser(pydantic_object=Joke)\n",
|
||||
"\n",
|
||||
"prompt = PromptTemplate(\n",
|
||||
" template=\"Answer the user query.\\n{format_instructions}\\n{query}\\n\",\n",
|
||||
" input_variables=[\"query\"],\n",
|
||||
" partial_variables={\"format_instructions\": parser.get_format_instructions()}\n",
|
||||
")\n",
|
||||
"\n",
|
||||
"# And a query intended to prompt a language model to populate the data structure.\n",
|
||||
"prompt_and_model = prompt | model\n",
|
||||
"output = prompt_and_model.invoke({\"query\": \"Tell me a joke.\"})\n",
|
||||
"parser.invoke(output)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "75976cd6-78e2-458b-821f-3ddf3683466b",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"## LCEL\n",
|
||||
"\n",
|
||||
"Output parsers implement the [Runnable interface](/docs/expression_language/interface), the basic building block of the [LangChain Expression Language (LCEL)](/docs/expression_language/). This means they support `invoke`, `ainvoke`, `stream`, `astream`, `batch`, `abatch`, `astream_log` calls.\n",
|
||||
"\n",
|
||||
"Output parsers accept a string or `BaseMessage` as input and can return an arbitrary type."
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 7,
|
||||
"id": "34f7ff0c-8443-4eb9-8704-b4f821811d93",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"Joke(setup='Why did the chicken cross the road?', punchline='To get to the other side!')"
|
||||
]
|
||||
},
|
||||
"execution_count": 7,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"parser.invoke(output)"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "bdebf4a5-57a8-4632-bd17-56723d431cf1",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"Instead of manually invoking the parser, we also could've just added it to our `Runnable` sequence:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 8,
|
||||
"id": "51f7fff5-e9bd-49a1-b5ab-b9ff281b93cb",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"Joke(setup='Why did the chicken cross the road?', punchline='To get to the other side!')"
|
||||
]
|
||||
},
|
||||
"execution_count": 8,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"chain = prompt | model | parser\n",
|
||||
"chain.invoke({\"query\": \"Tell me a joke.\"})"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "d88590a0-f36b-4ad5-8a56-d300971a6440",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"While all parsers support the streaming interface, only certain parsers can stream through partially parsed objects, since this is highly dependent on the output type. Parsers which cannot construct partial objects will simply yield the fully parsed output.\n",
|
||||
"\n",
|
||||
"The `SimpleJsonOutputParser` for example can stream through partial outputs:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 16,
|
||||
"id": "d7ecfe4d-dae8-4452-98ea-e48bdc498788",
|
||||
"metadata": {},
|
||||
"outputs": [],
|
||||
"source": [
|
||||
"from langchain.output_parsers.json import SimpleJsonOutputParser\n",
|
||||
"\n",
|
||||
"json_prompt = PromptTemplate.from_template(\"Return a JSON object with an `answer` key that answers the following question: {question}\")\n",
|
||||
"json_parser = SimpleJsonOutputParser()\n",
|
||||
"json_chain = json_prompt | model | json_parser"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 17,
|
||||
"id": "cc2b999e-47aa-41f4-ba6a-13b20a204576",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"[{},\n",
|
||||
" {'answer': ''},\n",
|
||||
" {'answer': 'Ant'},\n",
|
||||
" {'answer': 'Anton'},\n",
|
||||
" {'answer': 'Antonie'},\n",
|
||||
" {'answer': 'Antonie van'},\n",
|
||||
" {'answer': 'Antonie van Lee'},\n",
|
||||
" {'answer': 'Antonie van Leeu'},\n",
|
||||
" {'answer': 'Antonie van Leeuwen'},\n",
|
||||
" {'answer': 'Antonie van Leeuwenho'},\n",
|
||||
" {'answer': 'Antonie van Leeuwenhoek'}]"
|
||||
]
|
||||
},
|
||||
"execution_count": 17,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"list(json_chain.stream({\"question\": \"Who invented the microscope?\"}))"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"id": "3ca23082-c602-4ee8-af8c-a185b1f42bd1",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
"While the PydanticOutputParser cannot:"
|
||||
]
|
||||
},
|
||||
{
|
||||
"cell_type": "code",
|
||||
"execution_count": 18,
|
||||
"id": "07420e8f-e144-42aa-93ac-de890b6222f5",
|
||||
"metadata": {},
|
||||
"outputs": [
|
||||
{
|
||||
"data": {
|
||||
"text/plain": [
|
||||
"[Joke(setup='Why did the chicken cross the road?', punchline='To get to the other side!')]"
|
||||
]
|
||||
},
|
||||
"execution_count": 18,
|
||||
"metadata": {},
|
||||
"output_type": "execute_result"
|
||||
}
|
||||
],
|
||||
"source": [
|
||||
"list(chain.stream({\"query\": \"Tell me a joke.\"}))"
|
||||
]
|
||||
}
|
||||
],
|
||||
"metadata": {
|
||||
"kernelspec": {
|
||||
"display_name": "poetry-venv",
|
||||
"language": "python",
|
||||
"name": "poetry-venv"
|
||||
},
|
||||
"language_info": {
|
||||
"codemirror_mode": {
|
||||
"name": "ipython",
|
||||
"version": 3
|
||||
},
|
||||
"file_extension": ".py",
|
||||
"mimetype": "text/x-python",
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.9.1"
|
||||
}
|
||||
},
|
||||
"nbformat": 4,
|
||||
"nbformat_minor": 5
|
||||
}
|
||||
@ -1,91 +0,0 @@
|
||||
---
|
||||
sidebar_position: 2
|
||||
---
|
||||
# Output parsers
|
||||
|
||||
Language models output text. But many times you may want to get more structured information than just text back. This is where output parsers come in.
|
||||
|
||||
Output parsers are classes that help structure language model responses. There are two main methods an output parser must implement:
|
||||
|
||||
- "Get format instructions": A method which returns a string containing instructions for how the output of a language model should be formatted.
|
||||
- "Parse": A method which takes in a string (assumed to be the response from a language model) and parses it into some structure.
|
||||
|
||||
And then one optional one:
|
||||
|
||||
- "Parse with prompt": A method which takes in a string (assumed to be the response from a language model) and a prompt (assumed to be the prompt that generated such a response) and parses it into some structure. The prompt is largely provided in the event the OutputParser wants to retry or fix the output in some way, and needs information from the prompt to do so.
|
||||
|
||||
## Get started
|
||||
|
||||
Below we go over the main type of output parser, the `PydanticOutputParser`.
|
||||
|
||||
```python
|
||||
from langchain.prompts import PromptTemplate, ChatPromptTemplate, HumanMessagePromptTemplate
|
||||
from langchain.llms import OpenAI
|
||||
from langchain.chat_models import ChatOpenAI
|
||||
|
||||
from langchain.output_parsers import PydanticOutputParser
|
||||
from pydantic import BaseModel, Field, validator
|
||||
from typing import List
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
model_name = 'text-davinci-003'
|
||||
temperature = 0.0
|
||||
model = OpenAI(model_name=model_name, temperature=temperature)
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
# Define your desired data structure.
|
||||
class Joke(BaseModel):
|
||||
setup: str = Field(description="question to set up a joke")
|
||||
punchline: str = Field(description="answer to resolve the joke")
|
||||
|
||||
# You can add custom validation logic easily with Pydantic.
|
||||
@validator('setup')
|
||||
def question_ends_with_question_mark(cls, field):
|
||||
if field[-1] != '?':
|
||||
raise ValueError("Badly formed question!")
|
||||
return field
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
# Set up a parser + inject instructions into the prompt template.
|
||||
parser = PydanticOutputParser(pydantic_object=Joke)
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
prompt = PromptTemplate(
|
||||
template="Answer the user query.\n{format_instructions}\n{query}\n",
|
||||
input_variables=["query"],
|
||||
partial_variables={"format_instructions": parser.get_format_instructions()}
|
||||
)
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
# And a query intended to prompt a language model to populate the data structure.
|
||||
joke_query = "Tell me a joke."
|
||||
_input = prompt.format_prompt(query=joke_query)
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
output = model(_input.to_string())
|
||||
```
|
||||
|
||||
|
||||
```python
|
||||
parser.parse(output)
|
||||
```
|
||||
|
||||
<CodeOutputBlock lang="python">
|
||||
|
||||
```
|
||||
Joke(setup='Why did the chicken cross the road?', punchline='To get to the other side!')
|
||||
```
|
||||
|
||||
</CodeOutputBlock>
|
||||
Loading…
Reference in New Issue