community[patch]: Allow adding ARNs as model_id to support Amazon Bed…

…rock custom models (#16800)

- **Description:** Adds an additional class variable to `BedrockBase`
called `provider` that allows sending a model provider such as amazon,
cohere, ai21, etc.
Up until now, the model provider is extracted from the `model_id` using
the first part before the `.`, such as `amazon` for
`amazon.titan-text-express-v1` (see [supported list of Bedrock model IDs
here](https://docs.aws.amazon.com/bedrock/latest/userguide/model-ids-arns.html)).
But for custom Bedrock models where the ARN of the provisioned
throughput must be supplied, the `model_id` is like
`arn:aws:bedrock:...` so the `model_id` cannot be extracted from this. A
model `provider` is required by the LangChain Bedrock class to perform
model-based processing. To allow the same processing to be performed for
custom-models of a specific base model type, passing this `provider`
argument can help solve the issues.
The alternative considered here was the use of
`provider.arn:aws:bedrock:...` which then requires ARN to be extracted
and passed separately when invoking the model. The proposed solution
here is simpler and also does not cause issues for current models
already using the Bedrock class.
  - **Issue:** N/A
  - **Dependencies:** N/A

---------

Co-authored-by: Piyush Jain <[email protected]>

docs/docs/integrations/llms/bedrock.ipynb

@@ -111,7 +111,35 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-
+    "### Custom models"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "custom_llm = Bedrock(\n",
+    "    credentials_profile_name=\"bedrock-admin\",\n",
+    "    provider=\"cohere\",\n",
+    "    model_id=\"<Custom model ARN>\",  # ARN like 'arn:aws:bedrock:...' obtained via provisioning the custom model\n",
+    "    model_kwargs={\"temperature\": 1},\n",
+    "    streaming=True,\n",
+    "    callbacks=[StreamingStdOutCallbackHandler()],\n",
+    ")\n",
+    "\n",
+    "conversation = ConversationChain(\n",
+    "    llm=custom_llm, verbose=True, memory=ConversationBufferMemory()\n",
+    ")\n",
+    "conversation.predict(input=\"What is the recipe of mayonnaise?\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Guardrails for Amazon Bedrock example \n",
     "\n",
     "## Guardrails for Amazon Bedrock (Preview) \n",
     "[Guardrails for Amazon Bedrock](https://aws.amazon.com/bedrock/guardrails/) evaluates user inputs and model responses based on use case specific policies, and provides an additional layer of safeguards regardless of the underlying model. Guardrails can be applied across models, including Anthropic Claude, Meta Llama 2, Cohere Command, AI21 Labs Jurassic, and Amazon Titan Text, as well as fine-tuned models.\n",
@@ -139,7 +167,7 @@
     "            print(f\"Guardrails: {kwargs}\")\n",
     "\n",
     "\n",
-    "# guardrails for Amazon Bedrock with trace\n",
+    "# Guardrails for Amazon Bedrock with trace\n",
     "llm = Bedrock(\n",
     "    credentials_profile_name=\"bedrock-admin\",\n",
     "    model_id=\"<Model_ID>\",\n",
@@ -166,7 +194,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.12"
+   "version": "3.11.7"
   }
  },
  "nbformat": 4,

libs/community/langchain_community/llms/bedrock.py

@@ -229,9 +229,17 @@ class BedrockBase(BaseModel, ABC):
     config: Optional[Config] = None
     """An optional botocore.config.Config instance to pass to the client."""
 
+    provider: Optional[str] = None
+    """The model provider, e.g., amazon, cohere, ai21, etc. When not supplied, provider
+    is extracted from the first part of the model_id e.g. 'amazon' in 
+    'amazon.titan-text-express-v1'. This value should be provided for model ids that do
+    not have the provider in them, e.g., custom and provisioned models that have an ARN
+    associated with them."""
+
     model_id: str
     """Id of the model to call, e.g., amazon.titan-text-express-v1, this is
-    equivalent to the modelId property in the list-foundation-models api"""
+    equivalent to the modelId property in the list-foundation-models api. For custom and
+    provisioned models, an ARN value is expected."""
 
     model_kwargs: Optional[Dict] = None
     """Keyword arguments to pass to the model."""
@@ -353,6 +361,14 @@ def _identifying_params(self) -> Mapping[str, Any]:
         }
 
     def _get_provider(self) -> str:
+        if self.provider:
+            return self.provider
+        if self.model_id.startswith("arn"):
+            raise ValueError(
+                "Model provider should be supplied when passing a model ARN as "
+                "model_id"
+            )
+
         return self.model_id.split(".")[0]
 
     @property