Merge branch 'dev' into refactor/loaders

.github/dependabot.yml

@@ -15,6 +15,9 @@ updates:
         update-types:
         - "minor"
         - "patch"
+    allow:
+      - dependency-type: production
+      - dependency-type: development
   - package-ecosystem: "github-actions"
     directory: "/"
     schedule:

CHANGELOG.md

@@ -6,25 +6,36 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## Unreleased
 ### Added
-- `BaseFileLoader` for Loaders that load from a path.
-- `BaseLoader.fetch()` method for fetching data from a source.
-- `BaseLoader.parse()` method for parsing fetched data.
-- `BaseFileManager.encoding` to specify the encoding when loading and saving files.
-- `BaseWebScraperDriver.extract_page()` method for extracting data from an already scraped web page.
-- `TextLoaderRetrievalRagModule.chunker` for specifying the chunking strategy.
-- `file_utils.get_mime_type` utility for getting the MIME type of a file.
+- `Workflow.input_tasks` and `Workflow.output_tasks` to access the input and output tasks of a Workflow.
+- Ability to pass nested list of `Tasks` to `Structure.tasks` allowing for more complex declarative Structure definitions.
+
+
+## Added
+- Parameter `pipeline_task` on `HuggingFacePipelinePromptDriver` for creating different types of `Pipeline`s.
+- `TavilyWebSearchDriver` to integrate Tavily's web search capabilities.
 
 ### Changed
-- **BREAKING**: Removed `BaseFileManager.default_loader` and `BaseFileManager.loaders`.
-- **BREAKING**: Loaders no longer chunk data, use a Chunker to chunk the data.
-- **BREAKING**: Removed `fileutils.load_file` and `fileutils.load_files`.
-- **BREAKING**: Removed `loaders-dataframe` and `loaders-audio` extras as they are no longer needed.
-- **BREKING**: `TextLoader`, `PdfLoader`, `ImageLoader`, and `AudioLoader` now take a `str | PathLike` instead of `bytes`.
-- **BREAKING**: Removed `DataframeLoader`.
-- `LocalFileManagerDriver.workdir` is now optional.
-- `filetype` is now a core dependency.
-- `FileManagerTool` now uses `filetype` for more accurate file type detection.
-- `BaseFileLoader.load_file()` will now either return a `TextArtifact` or a `BlobArtifact` depending on whether `BaseFileManager.encoding` is set.
+- **BREAKING**: Renamed parameters on several classes to `client`:
+  - `bedrock_client` on `AmazonBedrockCohereEmbeddingDriver`.
+  - `bedrock_client` on `AmazonBedrockCohereEmbeddingDriver`.
+  - `bedrock_client` on `AmazonBedrockTitanEmbeddingDriver`.
+  - `bedrock_client` on `AmazonBedrockImageGenerationDriver`.
+  - `bedrock_client` on `AmazonBedrockImageQueryDriver`.
+  - `bedrock_client` on `AmazonBedrockPromptDriver`.
+  - `sagemaker_client` on `AmazonSageMakerJumpstartEmbeddingDriver`.
+  - `sagemaker_client` on `AmazonSageMakerJumpstartPromptDriver`.
+  - `sqs_client` on `AmazonSqsEventListenerDriver`.
+  - `iotdata_client` on `AwsIotCoreEventListenerDriver`.
+  - `s3_client` on `AmazonS3FileManagerDriver`.
+  - `s3_client` on `AwsS3Tool`.
+  - `iam_client` on `AwsIamTool`.
+  - `pusher_client` on `PusherEventListenerDriver`.
+  - `mq` on `MarqoVectorStoreDriver`.
+  - `model_client` on `GooglePromptDriver`.
+  - `model_client` on `GoogleTokenizer`.
+- **BREAKING**: Renamed parameter `pipe` on `HuggingFacePipelinePromptDriver` to `pipeline`.
+- Several places where API clients are initialized are now lazy loaded.
+
 
 ## [0.32.0] - 2024-09-17
 

docs/griptape-cloud/data-sources/create-data-source.md

@@ -10,6 +10,10 @@ You can [create a Data Source in the Griptape Cloud console](https://cloud.gript
 
 You can scrape and ingest a single, public web page by providing a URL. If you wish to scrape multiple pages, you must create multiple Data Sources. However, you can then add all of the pages to the same Knowledge Base if you wish to access all the pages together.
 
+### Amazon S3
+
+You can connect Amazon S3 buckets, objects, and prefixes by providing their S3 URI(s). Supported file extensions include .pdf, .csv, .md, and most text-based file types.
+
 ### Google Drive
 
 You can ingest documents and spreadsheets stored in a Google Drive account. We support all standard file formats such as text, markdown, spreadsheets, and presentations.

docs/griptape-framework/drivers/src/web_search_drivers_4.py

@@ -0,0 +1,7 @@
+import os
+
+from griptape.drivers import TavilyWebSearchDriver
+
+driver = TavilyWebSearchDriver(api_key=os.environ["TAVILY_API_KEY"])
+
+driver.search("griptape ai")

docs/griptape-framework/drivers/src/web_search_drivers_5.py

@@ -0,0 +1,9 @@
+from griptape.drivers import DuckDuckGoWebSearchDriver
+from griptape.structures import Agent
+from griptape.tools import PromptSummaryTool, WebSearchTool
+
+agent = Agent(
+    tools=[WebSearchTool(web_search_driver=DuckDuckGoWebSearchDriver()), PromptSummaryTool(off_prompt=False)],
+)
+
+agent.run("Give me some websites with information about AI frameworks.")

docs/griptape-framework/drivers/web-search-drivers.md

@@ -1,6 +1,6 @@
 ---
 search:
-  boost: 2 
+  boost: 2
 ---
 
 ## Overview
@@ -9,7 +9,47 @@ Web Search Drivers can be used to search for links from a search query. They are
 
 * `search()` searches the web and returns a [ListArtifact](../../reference/griptape/artifacts/list_artifact.md) that contains JSON-serializable [TextArtifact](../../reference/griptape/artifacts/text_artifact.md)s with the search results.
 
-## Vector Store Drivers
+You can use Web Search Drivers with [Structures](../structures/agents.md):
+
+```python
+--8<-- "docs/griptape-framework/drivers/src/web_search_drivers_5.py"
+```
+```
+ToolkitTask 45a53f1024494baab41a1f10a67017b1
+    Output: Here are some websites with information about AI
+    frameworks:
+
+      1. [The Top 16 AI Frameworks and Libraries: A Beginner's Guide -
+      DataCamp](https://www.datacamp.com/blog/top-ai-frameworks-and-lib
+      raries)
+      2. [AI Frameworks: Top Types To Adopt in 2024 -
+      Splunk](https://www.splunk.com/en_us/blog/learn/ai-frameworks.htm
+      l)
+      3. [Top AI Frameworks in 2024: A Review -
+      BairesDev](https://www.bairesdev.com/blog/ai-frameworks/)
+      4. [The Top 16 AI Frameworks and Libraries - AI
+      Slackers](https://aislackers.com/the-top-16-ai-frameworks-and-lib
+      raries/)
+      5. [Top AI Frameworks in 2024: Artificial Intelligence Frameworks
+      Comparison - Clockwise
+      Software](https://clockwise.software/blog/artificial-intelligence
+      -framework/)
+```
+Or use them independently:
+
+```python
+--8<-- "docs/griptape-framework/drivers/src/web_search_drivers_3.py"
+```
+```
+{"title": "The Top 16 AI Frameworks and Libraries: A Beginner's Guide", "url": "https://www.datacamp.com/blog/top-ai-frameworks-and-libraries", "description": "PyTorch. Torch is an open-source machine learning library known for its dynamic computational graph and is favored by researchers. The framework is excellent for prototyping and experimentation. Moreover, it's empowered by growing community support, with tools like PyTorch being built on the library."}
+
+{"title": "Top 11 AI Frameworks and Tools in 2024 | Fively | 5ly.co", "url": "https://5ly.co/blog/best-ai-frameworks/", "description": "Discover the top 11 modern artificial intelligence tools and frameworks to build robust architectures for your AI-powered apps. ... - Some advanced use cases may need further fine-tuning. Caffe 2. Now we move on to deep learning tools and frameworks. The first one is Caffe 2: an open-source deep learning framework with modularity and speed in ..."}
+
+{"title": "The Top 16 AI Frameworks and Libraries | AI Slackers", "url": "https://aislackers.com/the-top-16-ai-frameworks-and-libraries/", "description": "Experiment with different frameworks to find the one that aligns with your needs and goals as a data practitioner. Embrace the world of AI frameworks, and embark on a journey of building smarter software with confidence. Discover the top AI frameworks and libraries like PyTorch, Scikit-Learn, TensorFlow, Keras, LangChain, and more."}
+```
+
+
+## Web Search Drivers
 
 ### Google
 
@@ -21,12 +61,6 @@ Example using `GoogleWebSearchDriver` directly:
 --8<-- "docs/griptape-framework/drivers/src/web_search_drivers_1.py"
 ```
 
-Example of using `GoogleWebSearchDriver` with an agent:
-
-```python
---8<-- "docs/griptape-framework/drivers/src/web_search_drivers_2.py"
-```
-
 ### DuckDuckGo
 
 !!! info
@@ -39,3 +73,13 @@ Example of using `DuckDuckGoWebSearchDriver` directly:
 ```python
 --8<-- "docs/griptape-framework/drivers/src/web_search_drivers_3.py"
 ```
+
+### Tavily
+!!! info
+    This driver requires the `drivers-web-search-tavily` [extra](../index.md#extras), and a Tavily [api key](https://app.tavily.com).
+
+Example of using `TavilyWebSearchDriver` directly:
+
+```python
+--8<-- "docs/griptape-framework/drivers/src/web_search_drivers_4.py"
+```

griptape/drivers/__init__.py

@@ -99,6 +99,7 @@
 from .web_search.base_web_search_driver import BaseWebSearchDriver
 from .web_search.google_web_search_driver import GoogleWebSearchDriver
 from .web_search.duck_duck_go_web_search_driver import DuckDuckGoWebSearchDriver
+from .web_search.tavily_web_search_driver import TavilyWebSearchDriver
 
 from .event_listener.base_event_listener_driver import BaseEventListenerDriver
 from .event_listener.amazon_sqs_event_listener_driver import AmazonSqsEventListenerDriver
@@ -213,6 +214,7 @@
     "BaseWebSearchDriver",
     "GoogleWebSearchDriver",
     "DuckDuckGoWebSearchDriver",
+    "TavilyWebSearchDriver",
     "BaseEventListenerDriver",
     "AmazonSqsEventListenerDriver",
     "WebhookEventListenerDriver",

griptape/drivers/audio_transcription/openai_audio_transcription_driver.py

@@ -4,10 +4,11 @@
 from typing import Optional
 
 import openai
-from attrs import Factory, define, field
+from attrs import define, field
 
 from griptape.artifacts import AudioArtifact, TextArtifact
 from griptape.drivers import BaseAudioTranscriptionDriver
+from griptape.utils.decorators import lazy_property
 
 
 @define
@@ -17,12 +18,11 @@ class OpenAiAudioTranscriptionDriver(BaseAudioTranscriptionDriver):
     base_url: Optional[str] = field(default=None, kw_only=True, metadata={"serializable": True})
     api_key: Optional[str] = field(default=None, kw_only=True, metadata={"serializable": False})
     organization: Optional[str] = field(default=openai.organization, kw_only=True, metadata={"serializable": True})
-    client: openai.OpenAI = field(
-        default=Factory(
-            lambda self: openai.OpenAI(api_key=self.api_key, base_url=self.base_url, organization=self.organization),
-            takes_self=True,
-        ),
-    )
+    _client: openai.OpenAI = field(default=None, kw_only=True, alias="client", metadata={"serializable": False})
+
+    @lazy_property()
+    def client(self) -> openai.OpenAI:
+        return openai.OpenAI(api_key=self.api_key, base_url=self.base_url, organization=self.organization)
 
     def try_run(self, audio: AudioArtifact, prompts: Optional[list[str]] = None) -> TextArtifact:
         additional_params = {}

griptape/drivers/embedding/amazon_bedrock_cohere_embedding_driver.py

@@ -1,16 +1,18 @@
 from __future__ import annotations
 
 import json
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING
 
 from attrs import Factory, define, field
 
 from griptape.drivers import BaseEmbeddingDriver
 from griptape.tokenizers.amazon_bedrock_tokenizer import AmazonBedrockTokenizer
 from griptape.utils import import_optional_dependency
+from griptape.utils.decorators import lazy_property
 
 if TYPE_CHECKING:
     import boto3
+    from mypy_boto3_bedrock import BedrockClient
 
     from griptape.tokenizers.base_tokenizer import BaseTokenizer
 
@@ -26,7 +28,7 @@ class AmazonBedrockCohereEmbeddingDriver(BaseEmbeddingDriver):
             `search_query` when querying your vector DB to find relevant documents.
         session: Optionally provide custom `boto3.Session`.
         tokenizer: Optionally provide custom `BedrockCohereTokenizer`.
-        bedrock_client: Optionally provide custom `bedrock-runtime` client.
+        client: Optionally provide custom `bedrock-runtime` client.
     """
 
     DEFAULT_MODEL = "cohere.embed-english-v3"
@@ -38,15 +40,16 @@ class AmazonBedrockCohereEmbeddingDriver(BaseEmbeddingDriver):
         default=Factory(lambda self: AmazonBedrockTokenizer(model=self.model), takes_self=True),
         kw_only=True,
     )
-    bedrock_client: Any = field(
-        default=Factory(lambda self: self.session.client("bedrock-runtime"), takes_self=True),
-        kw_only=True,
-    )
+    _client: BedrockClient = field(default=None, kw_only=True, alias="client", metadata={"serializable": False})
+
+    @lazy_property()
+    def client(self) -> BedrockClient:
+        return self.session.client("bedrock-runtime")
 
     def try_embed_chunk(self, chunk: str) -> list[float]:
         payload = {"input_type": self.input_type, "texts": [chunk]}
 
-        response = self.bedrock_client.invoke_model(
+        response = self.client.invoke_model(
             body=json.dumps(payload),
             modelId=self.model,
             accept="*/*",

griptape/drivers/embedding/amazon_bedrock_titan_embedding_driver.py

@@ -1,16 +1,18 @@
 from __future__ import annotations
 
 import json
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING
 
 from attrs import Factory, define, field
 
 from griptape.drivers import BaseEmbeddingDriver
 from griptape.tokenizers.amazon_bedrock_tokenizer import AmazonBedrockTokenizer
 from griptape.utils import import_optional_dependency
+from griptape.utils.decorators import lazy_property
 
 if TYPE_CHECKING:
     import boto3
+    from mypy_boto3_bedrock import BedrockClient
 
     from griptape.tokenizers.base_tokenizer import BaseTokenizer
 
@@ -23,7 +25,7 @@ class AmazonBedrockTitanEmbeddingDriver(BaseEmbeddingDriver):
         model: Embedding model name. Defaults to DEFAULT_MODEL.
         tokenizer: Optionally provide custom `BedrockTitanTokenizer`.
         session: Optionally provide custom `boto3.Session`.
-        bedrock_client: Optionally provide custom `bedrock-runtime` client.
+        client: Optionally provide custom `bedrock-runtime` client.
     """
 
     DEFAULT_MODEL = "amazon.titan-embed-text-v1"
@@ -34,15 +36,16 @@ class AmazonBedrockTitanEmbeddingDriver(BaseEmbeddingDriver):
         default=Factory(lambda self: AmazonBedrockTokenizer(model=self.model), takes_self=True),
         kw_only=True,
     )
-    bedrock_client: Any = field(
-        default=Factory(lambda self: self.session.client("bedrock-runtime"), takes_self=True),
-        kw_only=True,
-    )
+    _client: BedrockClient = field(default=None, kw_only=True, alias="client", metadata={"serializable": False})
+
+    @lazy_property()
+    def client(self) -> BedrockClient:
+        return self.session.client("bedrock-runtime")
 
     def try_embed_chunk(self, chunk: str) -> list[float]:
         payload = {"inputText": chunk}
 
-        response = self.bedrock_client.invoke_model(
+        response = self.client.invoke_model(
             body=json.dumps(payload),
             modelId=self.model,
             accept="application/json",

griptape/drivers/embedding/amazon_sagemaker_jumpstart_embedding_driver.py

@@ -1,32 +1,35 @@
 from __future__ import annotations
 
 import json
-from typing import TYPE_CHECKING, Any, Optional
+from typing import TYPE_CHECKING, Optional
 
 from attrs import Factory, define, field
 
 from griptape.drivers import BaseEmbeddingDriver
 from griptape.utils import import_optional_dependency
+from griptape.utils.decorators import lazy_property
 
 if TYPE_CHECKING:
     import boto3
+    from mypy_boto3_sagemaker import SageMakerClient
 
 
 @define
 class AmazonSageMakerJumpstartEmbeddingDriver(BaseEmbeddingDriver):
     session: boto3.Session = field(default=Factory(lambda: import_optional_dependency("boto3").Session()), kw_only=True)
-    sagemaker_client: Any = field(
-        default=Factory(lambda self: self.session.client("sagemaker-runtime"), takes_self=True),
-        kw_only=True,
-    )
     endpoint: str = field(kw_only=True, metadata={"serializable": True})
     custom_attributes: str = field(default="accept_eula=true", kw_only=True, metadata={"serializable": True})
     inference_component_name: Optional[str] = field(default=None, kw_only=True, metadata={"serializable": True})
+    _client: SageMakerClient = field(default=None, kw_only=True, alias="client", metadata={"serializable": False})
+
+    @lazy_property()
+    def client(self) -> SageMakerClient:
+        return self.session.client("sagemaker-runtime")
 
     def try_embed_chunk(self, chunk: str) -> list[float]:
         payload = {"text_inputs": chunk, "mode": "embedding"}
 
-        endpoint_response = self.sagemaker_client.invoke_endpoint(
+        endpoint_response = self.client.invoke_endpoint(
             EndpointName=self.endpoint,
             ContentType="application/json",
             Body=json.dumps(payload).encode("utf-8"),