Skip to content

Commit

Permalink
Add suggested changes to guide and class code
Browse files Browse the repository at this point in the history
  • Loading branch information
@Sheepsta300
Sheepsta300 committed Dec 10, 2024
1 parent 4915fe0 commit 71ae221
Show file tree
Hide file tree
Showing 2 changed files with 50 additions and 117 deletions.
163 changes: 48 additions & 115 deletions docs/docs/integrations/tools/azure_content_safety.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -64,42 +64,33 @@
"outputs": [],
"source": [
"import os\n",
"\n",
"import getpass\n",
"from langchain import hub"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We will use a prompt to instruct the model. LangChain prompts can be configured, but for simplicity, we will use a premade prompt from LangSmith. This requires an API key, which can be set up [here](https://www.langchain.com/langsmith) after registration."
"Now we can use the `AzureContentSafetyTextTool` combined with a model, using `create_react_agent`."
]
},
{
"cell_type": "code",
"execution_count": 2,
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"LANGSMITH_KEY = os.environ[\"LANGSMITH_KEY\"]\n",
"prompt = hub.pull(\"hwchase17/structured-chat-agent\", api_key=LANGSMITH_KEY)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can use the `AzureContentSafetyTextTool` combined with a model, using `create_structured_chat_agent`."
"from langgraph.prebuilt import create_react_agent"
]
},
{
"cell_type": "code",
"execution_count": 5,
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
"from langchain.agents import AgentExecutor, create_structured_chat_agent\n",
"from langchain_community.tools.azure_ai_services.content_safety import (\n",
"from libs.community.langchain_community.tools.azure_ai_services.content_safety import (\n",
" AzureContentSafetyTextTool,\n",
")\n",
"from langchain_openai import AzureChatOpenAI"
Expand All @@ -121,12 +112,22 @@
},
{
"cell_type": "code",
"execution_count": 7,
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"os.environ[\"CONTENT_SAFETY_ENDPOINT\"] = getpass.getpass()\n",
"os.environ[\"CONTENT_SAFETY_KEY\"] = getpass.getpass()"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"content_endpoint = os.environ[\"CONTENT_SAFETY_ENDPOINT\"]\n",
"content_key = os.environ[\"CONTENT_SAFETY_KEY\"]"
"content_key = os.environ[\"CONTENT_SAFETY_API_KEY\"]"
]
},
{
Expand All @@ -145,37 +146,49 @@
},
{
"cell_type": "code",
"execution_count": 8,
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"cs = AzureContentSafetyTextTool(\n",
" content_safety_key=content_key,\n",
" content_safety_endpoint=content_endpoint,\n",
")"
"os.environ[\"AZURE_OPENAI_API_KEY\"] = getpass.getpass()\n",
"os.environ[\"OPENAI_API_VERSION\"] = getpass.getpass()\n",
"os.environ[\"GPT_MODEL\"] = getpass.getpass()\n",
"os.environ[\"AZURE_OPENAI_ENDPOINT\"] = getpass.getpass()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"tools = [cs]\n",
"\n",
"model = AzureChatOpenAI(\n",
" openai_api_version=os.environ[\"OPENAI_API_VERSION\"],\n",
" azure_deployment=os.environ[\"COMPLETIONS_MODEL\"],\n",
" azure_deployment=os.environ[\"GPT_MODEL\"],\n",
" azure_endpoint=os.environ[\"AZURE_OPENAI_ENDPOINT\"],\n",
" api_key=os.environ[\"AZURE_OPENAI_API_KEY\"],\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"cs = AzureContentSafetyTextTool(\n",
" content_safety_key=content_key,\n",
" content_safety_endpoint=content_endpoint,\n",
")\n",
"\n",
"tools = [cs]"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Creating an `AgentExecutor` chain allows a model to use tools to assist in it's response."
"Create a react agent to invoke the tool. "
]
},
{
Expand All @@ -184,11 +197,7 @@
"metadata": {},
"outputs": [],
"source": [
"agent = create_structured_chat_agent(model, tools, prompt)\n",
"\n",
"agent_executor = AgentExecutor(\n",
" agent=agent, tools=tools, verbose=True, handle_parsing_errors=True\n",
")"
"agent = create_react_agent(model, tools)"
]
},
{
Expand Down Expand Up @@ -230,20 +239,9 @@
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"'Hate: 2\\nSelfHarm: 0\\nSexual: 0\\nViolence: 0\\n'"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"cs.invoke({\"query\": input})"
]
Expand All @@ -264,76 +262,11 @@
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"\n",
"\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
"\u001b[32;1m\u001b[1;3m{\n",
" \"action\": \"azure_content_safety_tool\",\n",
" \"action_input\": {\n",
" \"query\": \"I hate you\"\n",
" }\n",
"}\u001b[0m\u001b[36;1m\u001b[1;3mHate: 2\n",
"SelfHarm: 0\n",
"Sexual: 0\n",
"Violence: 0\n",
"\u001b[0m\u001b[32;1m\u001b[1;3m{\n",
" \"action\": \"Final Answer\",\n",
" \"action_input\": \"The text contains hate sentiment with a severity level of 2. There is no indication of self-harm, sexual content, or violence.\"\n",
"}\u001b[0m\n",
"\n",
"\u001b[1m> Finished chain.\u001b[0m\n"
]
},
{
"data": {
"text/plain": [
"{'input': 'Can you check the following text for harmful content : I hate you',\n",
" 'output': 'The text contains hate sentiment with a severity level of 2. There is no indication of self-harm, sexual content, or violence.'}"
]
},
"execution_count": 17,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"agent_executor.invoke(\n",
" {\"input\": f\"Can you check the following text for harmful content : {input}\"}\n",
")"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Chaining"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"When creating an `AgentExecutor` as described earlier, an execution chain is formed. The sequence of events will be printed using the given prompt, and actions will occur in a chain-like manner."
]
},
{
"cell_type": "code",
"execution_count": 18,
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"agent = create_structured_chat_agent(model, tools, prompt)\n",
"\n",
"agent_executor = AgentExecutor(\n",
" agent=agent, tools=tools, verbose=True, handle_parsing_errors=True\n",
")"
"agent.invoke({\"messages\": [(\"user\", f\"Can you check the following text for harmful content : {input}\")]})"
]
},
{
Expand All @@ -347,7 +280,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"[Azure AI Content Safety Overview](https://learn.microsoft.com/azure/ai-services/content-safety/overview) | [Azure AI Content Safety Python SDK](https://learn.microsoft.com/python/api/overview/azure/ai-contentsafety-readme?view=azure-python)"
"[Azure AI Content Safety Overview](https://learn.microsoft.com/azure/ai-services/content-safety/overview) | [Azure AI Content Safety Python API](https://learn.microsoft.com/python/api/overview/azure/ai-contentsafety-readme?view=azure-python)"
]
}
],
Expand All @@ -367,7 +300,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.4"
"version": "3.12.8"
}
},
"nbformat": 4,
Expand Down
4 changes: 2 additions & 2 deletions libs/community/langchain_community/tools/azure_ai_services/content_safety.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -31,8 +31,8 @@ class AzureContentSafetyTextTool(BaseTool):
requests.
"""

content_safety_key: str = "" #: :meta private:
content_safety_endpoint: str = "" #: :meta private:
content_safety_key: Optional[str] = None #: :meta private:
content_safety_endpoint: Optional[str] = None #: :meta private:
content_safety_client: Any = None #: :meta private:

name: str = "azure_content_safety_tool"
Expand Down

0 comments on commit 71ae221

Please sign in to comment.