Docs: backend only mode (#4405)
* initial-cleanup-and-test * better-flow * config * Apply suggestions from code review Co-authored-by: KimberlyFields <46325568+KimberlyFields@users.noreply.github.com> * more-why-clarity * cleanup * code-review-and-tech-review --------- Co-authored-by: KimberlyFields <46325568+KimberlyFields@users.noreply.github.com>
This commit is contained in:
Mendon Kissling
GitHub
parent
19301b1cce
commit
1138869d0f
1 changed files with 71 additions and 102 deletions
|
|
@ -1,154 +1,123 @@
|
||||||
---
|
---
|
||||||
title: Backend-Only
|
title: Run Langflow in backend-only mode
|
||||||
sidebar_position: 4
|
sidebar_position: 4
|
||||||
slug: /configuration-backend-only
|
slug: /configuration-backend-only
|
||||||
---
|
---
|
||||||
|
|
||||||
|
Langflow can run in `--backend-only` mode to expose a Langflow app as an API endpoint, without running the frontend UI.
|
||||||
|
This is also known as "headless" mode. Running Langflow without the frontend is useful for automation, testing, and situations where you just need to serve a flow as a workload without creating a new flow in the UI.
|
||||||
|
|
||||||
|
To run Langflow in backend-only mode, pass the `--backend-only` flag at startup.
|
||||||
|
|
||||||
:::info
|
```python
|
||||||
|
python3 -m langflow run --backend-only
|
||||||
|
```
|
||||||
|
|
||||||
This page may contain outdated information. It will be updated as soon as possible.
|
The terminal prints `Welcome to ⛓ Langflow`, and Langflow will now serve requests to its API without the frontend running.
|
||||||
|
|
||||||
:::
|
## Set up a basic prompting flow in backend-only mode
|
||||||
|
|
||||||
|
This example shows you how to set up a [Basic Prompting flow](/starter-projects-basic-prompting) as an endpoint in backend-only mode.
|
||||||
|
However, you can use these same instructions as guidelines for using any type of flow in backend-only mode.
|
||||||
|
|
||||||
|
### Prerequisites
|
||||||
|
|
||||||
|
- [Langflow is installed](/getting-started-installation)
|
||||||
|
- [You have an OpenAI API key](https://platform.openai.com/)
|
||||||
|
- [You have a Langflow Basic Prompting flow](/starter-projects-basic-prompting)
|
||||||
|
|
||||||
You can run Langflow in `--backend-only` mode to expose your Langflow app as an API, without running the frontend UI.
|
### Get your flow's ID
|
||||||
|
|
||||||
|
This guide assumes you have created a [Basic Prompting flow](/starter-projects-basic-prompting) or have another working flow available.
|
||||||
|
|
||||||
Start langflow in backend-only mode with `python3 -m langflow run --backend-only`.
|
1. In the Langflow UI, click **API**.
|
||||||
|
2. Click **curl** > **Copy code** to copy the curl command.
|
||||||
|
This command will POST input to your flow's endpoint.
|
||||||
The terminal prints `Welcome to ⛓ Langflow`, and a blank window opens at `http://127.0.0.1:7864/all`.
|
|
||||||
Langflow will now serve requests to its API without the frontend running.
|
|
||||||
|
|
||||||
|
|
||||||
## Prerequisites {#81dfa9407ed648889081b9d08b0e5cfe}
|
|
||||||
|
|
||||||
- [Langflow installed](/getting-started-installation)
|
|
||||||
- [OpenAI API key](https://platform.openai.com/)
|
|
||||||
- [A Langflow flow created](/starter-projects-basic-prompting)
|
|
||||||
|
|
||||||
## Download your flow's curl call {#d2cf1b694e4741eca07fd9806516007b}
|
|
||||||
|
|
||||||
1. Click API.
|
|
||||||
2. Click **curl** > **Copy code** and save the code to your local machine.
|
|
||||||
It will look something like this:
|
It will look something like this:
|
||||||
|
|
||||||
```text
|
```text
|
||||||
curl -X POST \\
|
curl -X POST \
|
||||||
"<http://127.0.0.1:7864/api/v1/run/ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef?stream=false>" \\
|
"http://127.0.0.1:7861/api/v1/run/fff8dcaa-f0f6-4136-9df0-b7cb38de42e0?stream=false" \
|
||||||
-H 'Content-Type: application/json'\\
|
-H 'Content-Type: application/json'\
|
||||||
-d '{"input_value": "message",
|
-d '{"input_value": "message",
|
||||||
"output_type": "chat",
|
"output_type": "chat",
|
||||||
"input_type": "chat",
|
"input_type": "chat",
|
||||||
"tweaks": {
|
"tweaks": {
|
||||||
"Prompt-kvo86": {},
|
"ChatInput-8a86T": {},
|
||||||
"OpenAIModel-MilkD": {},
|
"Prompt-pKfl9": {},
|
||||||
"ChatOutput-ktwdw": {},
|
"ChatOutput-WcGpD": {},
|
||||||
"ChatInput-xXC4F": {}
|
"OpenAIModel-5UyvQ": {}
|
||||||
}}'
|
}}'
|
||||||
|
|
||||||
```
|
```
|
||||||
|
|
||||||
|
The flow ID in this example is `fff8dcaa-f0f6-4136-9df0-b7cb38de42e0`, a UUID generated by Langflow and used in the endpoint URL.
|
||||||
|
See [Project & General Settings](/settings-project-general-settings) to change the endpoint.
|
||||||
|
|
||||||
Note the flow ID of `ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef`. You can find this ID in the UI as well to ensure you're querying the right flow.
|
3. To stop Langflow, press **Ctrl+C**.
|
||||||
|
|
||||||
|
### Start Langflow in backend-only mode
|
||||||
|
|
||||||
## Start Langflow in backend-only mode {#f0ba018daf3041c39c0d226dadf78d35}
|
1. Start Langflow in backend-only mode.
|
||||||
|
|
||||||
1. Stop Langflow with Ctrl+C.
|
```python
|
||||||
2. Start langflow in backend-only mode with `python3 -m langflow run --backend-only`.
|
python3 -m langflow run --backend-only
|
||||||
The terminal prints `Welcome to ⛓ Langflow`, and a blank window opens at `http://127.0.0.1:7864/all`.
|
```
|
||||||
Langflow will now serve requests to its API.
|
|
||||||
3. Run the curl code you copied from the UI.
|
The terminal prints `Welcome to ⛓ Langflow`.
|
||||||
|
Langflow is now serving requests to its API.
|
||||||
|
|
||||||
|
2. Run the curl code you copied from the UI.
|
||||||
You should get a result like this:
|
You should get a result like this:
|
||||||
|
|
||||||
```shell
|
```shell
|
||||||
{"session_id":"ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef:bf81d898868ac87e1b4edbd96c131c5dee801ea2971122cc91352d144a45b880","outputs":[{"inputs":{"input_value":"hi, are you there?"},"outputs":[{"results":{"result":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?"},"artifacts":{"message":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?","sender":"Machine","sender_name":"AI"},"messages":[{"message":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?","sender":"Machine","sender_name":"AI","component_id":"ChatOutput-ktwdw"}],"component_display_name":"Chat Output","component_id":"ChatOutput-ktwdw","used_frozen_result":false}]}]}%
|
{"session_id":"ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef:bf81d898868ac87e1b4edbd96c131c5dee801ea2971122cc91352d144a45b880","outputs":[{"inputs":{"input_value":"hi, are you there?"},"outputs":[{"results":{"result":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?"},"artifacts":{"message":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?","sender":"Machine","sender_name":"AI"},"messages":[{"message":"Arrr, ahoy matey! Aye, I be here. What be ye needin', me hearty?","sender":"Machine","sender_name":"AI","component_id":"ChatOutput-ktwdw"}],"component_display_name":"Chat Output","component_id":"ChatOutput-ktwdw","used_frozen_result":false}]}]}%
|
||||||
|
|
||||||
```
|
```
|
||||||
|
|
||||||
|
This confirms Langflow is receiving your POST request, running the flow, and returning the result without running the frontend.
|
||||||
|
|
||||||
Again, note that the flow ID matches.
|
You can interact with this endpoint using the other options in the **API** menu, including the Python and Javascript APIs.
|
||||||
Langflow is receiving your POST request, running the flow, and returning the result, all without running the frontend. Cool!
|
|
||||||
|
|
||||||
|
### Query the Langflow endpoint with a Python script
|
||||||
|
|
||||||
## Download your flow's Python API call {#5923ff9dc40843c7a22a72fa6c66540c}
|
Using the same flow ID, run a Python sample script to send a query and get a prettified JSON response back.
|
||||||
|
|
||||||
|
1. Create a Python file and name it `langflow_api_demo.py`.
|
||||||
Instead of using curl, you can download your flow as a Python API call instead.
|
|
||||||
|
|
||||||
1. Click API.
|
|
||||||
2. Click **Python API** > **Copy code** and save the code to your local machine.
|
|
||||||
The code will look something like this:
|
|
||||||
|
|
||||||
```python
|
```python
|
||||||
import requests
|
import requests
|
||||||
from typing import Optional
|
import json
|
||||||
|
|
||||||
BASE_API_URL = "<http://127.0.0.1:7864/api/v1/run>"
|
def query_langflow(message):
|
||||||
FLOW_ID = "ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef"
|
url = "http://127.0.0.1:7861/api/v1/run/fff8dcaa-f0f6-4136-9df0-b7cb38de42e0"
|
||||||
# You can tweak the flow by adding a tweaks dictionary
|
headers = {"Content-Type": "application/json"}
|
||||||
# e.g {"OpenAI-XXXXX": {"model_name": "gpt-4"}}
|
data = {"input_value": message}
|
||||||
|
|
||||||
def run_flow(message: str,
|
response = requests.post(url, headers=headers, json=data)
|
||||||
flow_id: str,
|
|
||||||
output_type: str = "chat",
|
|
||||||
input_type: str = "chat",
|
|
||||||
tweaks: Optional[dict] = None,
|
|
||||||
api_key: Optional[str] = None) -> dict:
|
|
||||||
"""Run a flow with a given message and optional tweaks.
|
|
||||||
|
|
||||||
:param message: The message to send to the flow
|
|
||||||
:param flow_id: The ID of the flow to run
|
|
||||||
:param tweaks: Optional tweaks to customize the flow
|
|
||||||
:return: The JSON response from the flow
|
|
||||||
"""
|
|
||||||
api_url = f"{BASE_API_URL}/{flow_id}"
|
|
||||||
payload = {
|
|
||||||
"input_value": message,
|
|
||||||
"output_type": output_type,
|
|
||||||
"input_type": input_type,
|
|
||||||
}
|
|
||||||
headers = None
|
|
||||||
if tweaks:
|
|
||||||
payload["tweaks"] = tweaks
|
|
||||||
if api_key:
|
|
||||||
headers = {"x-api-key": api_key}
|
|
||||||
response = requests.post(api_url, json=payload, headers=headers)
|
|
||||||
return response.json()
|
return response.json()
|
||||||
|
|
||||||
# Setup any tweaks you want to apply to the flow
|
user_input = input("Enter your message: ")
|
||||||
|
result = query_langflow(user_input)
|
||||||
|
|
||||||
message = "message"
|
print(json.dumps(result, indent=2))
|
||||||
|
```
|
||||||
|
2. Run the script.
|
||||||
|
|
||||||
print(run_flow(message=message, flow_id=FLOW_ID))
|
```python
|
||||||
|
python langflow_api_demo.py
|
||||||
|
```
|
||||||
|
|
||||||
|
3. Enter your message when prompted.
|
||||||
|
You will get a prettified JSON response back containing a response to your message.
|
||||||
|
|
||||||
|
### Configure host and ports in backend-only mode
|
||||||
|
|
||||||
|
To change the host and port, pass the values as additional flags.
|
||||||
|
|
||||||
|
```python
|
||||||
|
python -m langflow run --host 127.0.0.1 --port 7860 --backend-only
|
||||||
```
|
```
|
||||||
|
|
||||||
|
|
||||||
3. Run your Python app:
|
|
||||||
|
|
||||||
|
|
||||||
```shell
|
|
||||||
python3 app.py
|
|
||||||
```
|
|
||||||
|
|
||||||
|
|
||||||
The result is similar to the curl call:
|
|
||||||
|
|
||||||
|
|
||||||
```json
|
|
||||||
{'session_id': 'ef7e0554-69e5-4e3e-ab29-ee83bcd8d9ef:bf81d898868ac87e1b4edbd96c131c5dee801ea2971122cc91352d144a45b880', 'outputs': [{'inputs': {'input_value': 'message'}, 'outputs': [{'results': {'result': "Arrr matey! What be yer message for this ol' pirate? Speak up or walk the plank!"}, 'artifacts': {'message': "Arrr matey! What be yer message for this ol' pirate? Speak up or walk the plank!", 'sender': 'Machine', 'sender_name': 'AI'}, 'messages': [{'message': "Arrr matey! What be yer message for this ol' pirate? Speak up or walk the plank!", 'sender': 'Machine', 'sender_name': 'AI', 'component_id': 'ChatOutput-ktwdw'}], 'component_display_name': 'Chat Output', 'component_id': 'ChatOutput-ktwdw', 'used_frozen_result': False}]}]}
|
|
||||||
|
|
||||||
```
|
|
||||||
|
|
||||||
|
|
||||||
Your Python app POSTs to your Langflow server, and the server runs the flow and returns the result.
|
|
||||||
|
|
||||||
|
|
||||||
See [API](https://www.notion.so/administration/api) for more ways to interact with your headless Langflow server.
|
|
||||||
|
|
||||||
|
|
|
||||||
Loading…
Add table
Add a link
Reference in a new issue