github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2026-01-25 14:35:49 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs: Integration Update - ADS4GPTs (#30153)

Browse Source 
docs: New integration for LangChain - ads4gpts-langchain

Description: Tools and Toolkit for Agentic integration natively within
LangChain with ADS4GPTs, in order to help applications monetize with
advertising.

Twitter handle: @ads4gpts

Co-authored-by: knitlydevaccount <loom+github@knitly.app>
This commit is contained in:
Ioannis Bakagiannis
2025-03-07 21:35:44 +02:00
committed by GitHub
parent 3c258194ae
commit 3444e587ee
3 changed files with 452 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

82
docs/docs/integrations/providers/ads4gpts.mdx Normal file
View File

@@ -0,0 +1,82 @@
# ADS4GPTs
> [ADS4GPTs](https://www.ads4gpts.com/) is building the open monetization backbone of the AI-Native internet. It helps AI applications monetize through advertising with a UX and Privacy first approach.
## Installation and Setup
### Using pip
You can install the package directly from PyPI:
```bash
pip install ads4gpts-langchain
```
### From Source
Alternatively, install from source:
```bash
git clone https://github.com/ADS4GPTs/ads4gpts.git
cd ads4gpts/libs/python-sdk/ads4gpts-langchain
pip install .
```
## Prerequisites
- Python 3.11+
- ADS4GPTs API Key ([Obtain API Key](https://www.ads4gpts.com))
## Environment Variables
Set the following environment variables for API authentication:
```bash
export ADS4GPTS_API_KEY='your-ads4gpts-api-key'
```
Alternatively, API keys can be passed directly when initializing classes or stored in a `.env` file.
## Tools
ADS4GPTs provides two main tools for monetization:
### Ads4gptsInlineSponsoredResponseTool
This tool fetches native, sponsored responses that can be seamlessly integrated within your AI application's outputs.
```python
from ads4gpts_langchain import Ads4gptsInlineSponsoredResponseTool
```
### Ads4gptsSuggestedPromptTool
Generates sponsored prompt suggestions to enhance user engagement and provide monetization opportunities.
```python
from ads4gpts_langchain import Ads4gptsSuggestedPromptTool
```
### Ads4gptsInlineConversationalTool
Delivers conversational sponsored content that naturally fits within chat interfaces and dialogs.
```python
from ads4gpts_langchain import Ads4gptsInlineConversationalTool
```
### Ads4gptsInlineBannerTool
Provides inline banner advertisements that can be displayed within your AI application's response.
```python
from ads4gpts_langchain import Ads4gptsInlineBannerTool
```
### Ads4gptsSuggestedBannerTool
Generates banner advertisement suggestions that can be presented to users as recommended content.
```python
from ads4gpts_langchain import Ads4gptsSuggestedBannerTool
```
## Toolkit
The `Ads4gptsToolkit` combines these tools for convenient access in LangChain applications.
```python
from ads4gpts_langchain import Ads4gptsToolkit
```

365
docs/docs/integrations/tools/ads4gpts.ipynb Normal file
View File

@@ -0,0 +1,365 @@
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# ADS4GPTs\n",
"\n",
"Integrate AI native advertising into your Agentic application.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Overview"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"This notebook outlines how to use the ADS4GPTs Tools and Toolkit in LangChain directly. In your LangGraph application though you will most likely use our prebuilt LangGraph agents."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Install ADS4GPTs Package\n",
"Install the ADS4GPTs package using pip."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Install ADS4GPTs Package\n",
"# Install the ADS4GPTs package using pip\n",
"!pip install ads4gpts-langchain"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Set up the environment variables for API authentication ([Obtain API Key](https://www.ads4gpts.com))."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Setup Environment Variables\n",
"# Prompt the user to enter their ADS4GPTs API key securely\n",
"if not os.environ.get(\"ADS4GPTS_API_KEY\"):\n",
" os.environ[\"ADS4GPTS_API_KEY\"] = getpass(\"Enter your ADS4GPTS API key: \")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Instantiation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Import the necessary libraries, including ADS4GPTs tools and toolkit.\n",
"\n",
"Initialize the ADS4GPTs tools such as Ads4gptsInlineSponsoredResponseTool. We are going to work with one tool because the process is the same for every other tool we provide."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Import Required Libraries\n",
"\n",
"import os\n",
"from getpass import getpass\n",
"\n",
"from ads4gpts_langchain import Ads4gptsInlineSponsoredResponseTool, Ads4gptsToolkit"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"# Initialize ADS4GPTs Tools\n",
"# Initialize the Ads4gptsInlineSponsoredResponseTool\n",
"inline_sponsored_response_tool = Ads4gptsInlineSponsoredResponseTool(\n",
" ads4gpts_api_key=os.environ[\"ADS4GPTS_API_KEY\"],\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Toolkit Instantiation\n",
"Initialize the Ads4gptsToolkit with the required parameters."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Initialized tool: Ads4gptsInlineSponsoredResponseTool\n",
"Initialized tool: Ads4gptsSuggestedPromptTool\n"
]
}
],
"source": [
"# Toolkit Initialization\n",
"# Initialize the Ads4gptsToolkit with the required parameters\n",
"toolkit = Ads4gptsToolkit(\n",
" ads4gpts_api_key=os.environ[\"ADS4GPTS_API_KEY\"],\n",
")\n",
"\n",
"# Retrieve tools from the toolkit\n",
"tools = toolkit.get_tools()\n",
"\n",
"# Print the initialized tools\n",
"for tool in tools:\n",
" print(f\"Initialized tool: {tool.__class__.__name__}\")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Invocation"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Run the ADS4GPTs tools with sample inputs and display the results."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Inline Sponsored Response Result: {'ad_text': '<- Promoted Content ->\\n\\nLearn the sartorial ways and get your handmade tailored suit by the masters themselves with Bespoke Tailors. [Subscribe now](https://youtube.com/@bespoketailorsdubai?si=9iH587ujoWKkueFa)\\n\\n<->'}\n"
]
}
],
"source": [
"# Run ADS4GPTs Tools\n",
"# Sample input data for the tools\n",
"sample_input = {\n",
" \"id\": \"test_id\",\n",
" \"user_gender\": \"female\",\n",
" \"user_age\": \"25-34\",\n",
" \"user_persona\": \"test_persona\",\n",
" \"ad_recommendation\": \"test_recommendation\",\n",
" \"undesired_ads\": \"test_undesired_ads\",\n",
" \"context\": \"test_context\",\n",
" \"num_ads\": 1,\n",
" \"style\": \"neutral\",\n",
"}\n",
"\n",
"# Run Ads4gptsInlineSponsoredResponseTool\n",
"inline_sponsored_response_result = inline_sponsored_response_tool._run(\n",
" **sample_input, ad_format=\"INLINE_SPONSORED_RESPONSE\"\n",
")\n",
"print(\"Inline Sponsored Response Result:\", inline_sponsored_response_result)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Async Run ADS4GPTs Tools\n",
"Run the ADS4GPTs tools asynchronously with sample inputs and display the results."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Async Inline Sponsored Response Result: {'ad_text': '<- Promoted Content ->\\n\\nGet the best tailoring content from Jonathan Farley. Learn to tie 100 knots and more! [Subscribe now](https://www.youtube.com/channel/UCx5hk4LN3p02jcUt3j_cexQ)\\n\\n<->'}\n"
]
}
],
"source": [
"import asyncio\n",
"\n",
"\n",
"# Define an async function to run the tools asynchronously\n",
"async def run_ads4gpts_tools_async():\n",
" # Run Ads4gptsInlineSponsoredResponseTool asynchronously\n",
" inline_sponsored_response_result = await inline_sponsored_response_tool._arun(\n",
" **sample_input, ad_format=\"INLINE_SPONSORED_RESPONSE\"\n",
" )\n",
" print(\"Async Inline Sponsored Response Result:\", inline_sponsored_response_result)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Toolkit Invocation\n",
"Use the Ads4gptsToolkit to get and run tools."
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Result from Ads4gptsInlineSponsoredResponseTool: {'ad_text': '<- Promoted Content ->\\n\\nLearn the sartorial ways and get your handmade tailored suit by the masters themselves with Bespoke Tailors. [Subscribe now](https://youtube.com/@bespoketailorsdubai?si=9iH587ujoWKkueFa)\\n\\n<->'}\n",
"Async result from Ads4gptsInlineSponsoredResponseTool: {'ad_text': '<- Promoted Content ->\\n\\nGet the best tailoring content from Jonathan Farley. Learn to tie 100 knots and more! [Subscribe now](https://www.youtube.com/channel/UCx5hk4LN3p02jcUt3j_cexQ)\\n\\n<->'}\n"
]
}
],
"source": [
"# Sample input data for the tools\n",
"sample_input = {\n",
" \"id\": \"test_id\",\n",
" \"user_gender\": \"female\",\n",
" \"user_age\": \"25-34\",\n",
" \"user_persona\": \"test_persona\",\n",
" \"ad_recommendation\": \"test_recommendation\",\n",
" \"undesired_ads\": \"test_undesired_ads\",\n",
" \"context\": \"test_context\",\n",
" \"num_ads\": 1,\n",
" \"style\": \"neutral\",\n",
"}\n",
"\n",
"# Run one tool and print the result\n",
"tool = tools[0]\n",
"result = tool._run(**sample_input)\n",
"print(f\"Result from {tool.__class__.__name__}:\", result)\n",
"\n",
"\n",
"# Define an async function to run the tools asynchronously\n",
"async def run_toolkit_tools_async():\n",
" result = await tool._arun(**sample_input)\n",
" print(f\"Async result from {tool.__class__.__name__}:\", result)\n",
"\n",
"\n",
"# Execute the async function\n",
"await run_toolkit_tools_async()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Chaining\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"if not os.environ.get(\"OPENAI_API_KEY\"):\n",
" os.environ[\"OPENAI_API_KEY\"] = getpass(\"Enter your OPENAI_API_KEY API key: \")"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Tool call: content='' additional_kwargs={'tool_calls': [{'id': 'call_XLR5UjF8JhylVHvrk9mTjhj8', 'function': {'arguments': '{\"id\":\"unique_user_id_001\",\"user_gender\":\"male\",\"user_age\":\"18-24\",\"ad_recommendation\":\"Stylish and trendy clothing suitable for young men going out with friends.\",\"undesired_ads\":\"formal wear, women\\'s clothing, children\\'s clothing\",\"context\":\"A young man looking for clothing to go out with friends\",\"num_ads\":1,\"style\":\"youthful and trendy\",\"ad_format\":\"INLINE_SPONSORED_RESPONSE\"}', 'name': 'ads4gpts_inline_sponsored_response'}, 'type': 'function'}], 'refusal': None} response_metadata={'token_usage': {'completion_tokens': 106, 'prompt_tokens': 1070, 'total_tokens': 1176, 'completion_tokens_details': {'accepted_prediction_tokens': 0, 'audio_tokens': 0, 'reasoning_tokens': 0, 'rejected_prediction_tokens': 0}, 'prompt_tokens_details': {'audio_tokens': 0, 'cached_tokens': 1024}}, 'model_name': 'gpt-4o-2024-08-06', 'system_fingerprint': 'fp_eb9dce56a8', 'finish_reason': 'tool_calls', 'logprobs': None} id='run-e3e64b4b-4505-4a71-bf02-a8d77bb68eee-0' tool_calls=[{'name': 'ads4gpts_inline_sponsored_response', 'args': {'id': 'unique_user_id_001', 'user_gender': 'male', 'user_age': '18-24', 'ad_recommendation': 'Stylish and trendy clothing suitable for young men going out with friends.', 'undesired_ads': \"formal wear, women's clothing, children's clothing\", 'context': 'A young man looking for clothing to go out with friends', 'num_ads': 1, 'style': 'youthful and trendy', 'ad_format': 'INLINE_SPONSORED_RESPONSE'}, 'id': 'call_XLR5UjF8JhylVHvrk9mTjhj8', 'type': 'tool_call'}] usage_metadata={'input_tokens': 1070, 'output_tokens': 106, 'total_tokens': 1176, 'input_token_details': {'audio': 0, 'cache_read': 1024}, 'output_token_details': {'audio': 0, 'reasoning': 0}}\n"
]
}
],
"source": [
"import os\n",
"\n",
"from langchain_openai import ChatOpenAI\n",
"\n",
"openai_model = ChatOpenAI(model=\"gpt-4o\", openai_api_key=os.environ[\"OPENAI_API_KEY\"])\n",
"model = openai_model.bind_tools(tools)\n",
"model_response = model.invoke(\n",
" \"Get me an ad for clothing. I am a young man looking to go out with friends.\"\n",
")\n",
"print(\"Tool call:\", model_response)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## API reference"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You can learn more about ADS4GPTs and the tools at our [GitHub](https://github.com/ADS4GPTs/ads4gpts/tree/main)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "ads4gpts-langraph-agent",
"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": "3.12.9"
}
},
"nbformat": 4,
"nbformat_minor": 2
}