not-again-ai 0.15.0__py3-none-any.whl → 0.16.0__py3-none-any.whl

not_again_ai/llm/embedding/__init__.py

@@ -0,0 +1,4 @@
+from not_again_ai.llm.embedding.interface import create_embeddings
+from not_again_ai.llm.embedding.types import EmbeddingRequest
+
+__all__ = ["EmbeddingRequest", "create_embeddings"]

not_again_ai/llm/embedding/interface.py

@@ -0,0 +1,28 @@
+from collections.abc import Callable
+from typing import Any
+
+from not_again_ai.llm.embedding.providers.ollama_api import ollama_create_embeddings
+from not_again_ai.llm.embedding.providers.openai_api import openai_create_embeddings
+from not_again_ai.llm.embedding.types import EmbeddingRequest, EmbeddingResponse
+
+
+def create_embeddings(request: EmbeddingRequest, provider: str, client: Callable[..., Any]) -> EmbeddingResponse:
+    """Get a embedding response from the given provider. Currently supported providers:
+    - `openai` - OpenAI
+    - `azure_openai` - Azure OpenAI
+    - `ollama` - Ollama
+
+    Args:
+        request: Request parameter object
+        provider: The supported provider name
+        client: Client information, see the provider's implementation for what can be provided
+
+    Returns:
+        EmbeddingResponse: The embedding response.
+    """
+    if provider == "openai" or provider == "azure_openai":
+        return openai_create_embeddings(request, client)
+    elif provider == "ollama":
+        return ollama_create_embeddings(request, client)
+    else:
+        raise ValueError(f"Provider {provider} not supported")

not_again_ai/llm/embedding/providers/ollama_api.py

@@ -0,0 +1,87 @@
+from collections.abc import Callable
+import os
+import re
+import time
+from typing import Any
+
+from loguru import logger
+from ollama import Client, EmbedResponse, ResponseError
+
+from not_again_ai.llm.embedding.types import EmbeddingObject, EmbeddingRequest, EmbeddingResponse
+
+OLLAMA_PARAMETER_MAP = {
+    "dimensions": None,
+}
+
+
+def validate(request: EmbeddingRequest) -> None:
+    # Check if any of the parameters set to OLLAMA_PARAMETER_MAP are not None
+    for key, value in OLLAMA_PARAMETER_MAP.items():
+        if value is None and getattr(request, key) is not None:
+            logger.warning(f"Parameter {key} is not supported by Ollama and will be ignored.")
+
+
+def ollama_create_embeddings(request: EmbeddingRequest, client: Callable[..., Any]) -> EmbeddingResponse:
+    validate(request)
+    kwargs = request.model_dump(mode="json", exclude_none=True)
+
+    # For each key in OLLAMA_PARAMETER_MAP
+    # If it is not None, set the key in kwargs to the value of the corresponding value in OLLAMA_PARAMETER_MAP
+    # If it is None, remove that key from kwargs
+    for key, value in OLLAMA_PARAMETER_MAP.items():
+        if value is not None and key in kwargs:
+            kwargs[value] = kwargs.pop(key)
+        elif value is None and key in kwargs:
+            del kwargs[key]
+
+    # Explicitly set truncate to True (it is the default)
+    kwargs["truncate"] = True
+
+    try:
+        start_time = time.time()
+        response: EmbedResponse = client(**kwargs)
+        end_time = time.time()
+        response_duration = round(end_time - start_time, 4)
+    except ResponseError as e:
+        # If the error says "model 'model' not found" use regex then raise a more specific error
+        expected_pattern = f"model '{request.model}' not found"
+        if re.search(expected_pattern, e.error):
+            raise ResponseError(f"Model '{request.model}' not found.") from e
+        else:
+            raise ResponseError(e.error) from e
+
+    embeddings: list[EmbeddingObject] = []
+    for index, embedding in enumerate(response.embeddings):
+        embeddings.append(EmbeddingObject(embedding=list(embedding), index=index))
+
+    return EmbeddingResponse(
+        embeddings=embeddings,
+        response_duration=response_duration,
+        total_tokens=response.prompt_eval_count,
+    )
+
+
+def ollama_client(host: str | None = None, timeout: float | None = None) -> Callable[..., Any]:
+    """Create an Ollama client instance based on the specified host or will read from the OLLAMA_HOST environment variable.
+
+    Args:
+        host (str, optional): The host URL of the Ollama server.
+        timeout (float, optional): The timeout for requests
+
+    Returns:
+        Client: An instance of the Ollama client.
+
+    Examples:
+        >>> client = client(host="http://localhost:11434")
+    """
+    if host is None:
+        host = os.getenv("OLLAMA_HOST")
+        if host is None:
+            logger.warning("OLLAMA_HOST environment variable not set, using default host: http://localhost:11434")
+            host = "http://localhost:11434"
+
+    def client_callable(**kwargs: Any) -> Any:
+        client = Client(host=host, timeout=timeout)
+        return client.embed(**kwargs)
+
+    return client_callable

not_again_ai/llm/embedding/providers/openai_api.py

@@ -0,0 +1,126 @@
+from collections.abc import Callable
+import time
+from typing import Any, Literal
+
+from azure.identity import DefaultAzureCredential, get_bearer_token_provider
+from openai import AzureOpenAI, OpenAI
+
+from not_again_ai.llm.embedding.types import EmbeddingObject, EmbeddingRequest, EmbeddingResponse
+
+
+def openai_create_embeddings(request: EmbeddingRequest, client: Callable[..., Any]) -> EmbeddingResponse:
+    kwargs = request.model_dump(mode="json", exclude_none=True)
+
+    start_time = time.time()
+    response = client(**kwargs)
+    end_time = time.time()
+    response_duration = round(end_time - start_time, 4)
+
+    embeddings: list[EmbeddingObject] = []
+    for data in response["data"]:
+        embeddings.append(EmbeddingObject(embedding=data["embedding"], index=data["index"]))
+
+    return EmbeddingResponse(
+        embeddings=embeddings,
+        response_duration=response_duration,
+        total_tokens=response["usage"]["total_tokens"],
+    )
+
+
+def create_client_callable(client_class: type[OpenAI | AzureOpenAI], **client_args: Any) -> Callable[..., Any]:
+    """Creates a callable that instantiates and uses an OpenAI client.
+
+    Args:
+        client_class: The OpenAI client class to instantiate (OpenAI or AzureOpenAI)
+        **client_args: Arguments to pass to the client constructor
+
+    Returns:
+        A callable that creates a client and returns completion results
+    """
+    filtered_args = {k: v for k, v in client_args.items() if v is not None}
+
+    def client_callable(**kwargs: Any) -> Any:
+        client = client_class(**filtered_args)
+        completion = client.embeddings.create(**kwargs)
+        return completion.to_dict()
+
+    return client_callable
+
+
+class InvalidOAIAPITypeError(Exception):
+    """Raised when an invalid OAIAPIType string is provided."""
+
+
+def openai_client(
+    api_type: Literal["openai", "azure_openai"] = "openai",
+    api_key: str | None = None,
+    organization: str | None = None,
+    aoai_api_version: str = "2024-06-01",
+    azure_endpoint: str | None = None,
+    timeout: float | None = None,
+    max_retries: int | None = None,
+) -> Callable[..., Any]:
+    """Create an OpenAI or Azure OpenAI client instance based on the specified API type and other provided parameters.
+
+    It is preferred to use RBAC authentication for Azure OpenAI. You must be signed in with the Azure CLI and have correct role assigned.
+    See https://techcommunity.microsoft.com/t5/microsoft-developer-community/using-keyless-authentication-with-azure-openai/ba-p/4111521
+
+    Args:
+        api_type (str, optional): Type of the API to be used. Accepted values are 'openai' or 'azure_openai'.
+            Defaults to 'openai'.
+        api_key (str, optional): The API key to authenticate the client. If not provided,
+            OpenAI automatically uses `OPENAI_API_KEY` from the environment.
+            If provided for Azure OpenAI, it will be used for authentication instead of the Azure AD token provider.
+        organization (str, optional): The ID of the organization. If not provided,
+            OpenAI automotically uses `OPENAI_ORG_ID` from the environment.
+        aoai_api_version (str, optional): Only applicable if using Azure OpenAI https://learn.microsoft.com/azure/ai-services/openai/reference#rest-api-versioning
+        azure_endpoint (str, optional): The endpoint to use for Azure OpenAI.
+        timeout (float, optional): By default requests time out after 10 minutes.
+        max_retries (int, optional): Certain errors are automatically retried 2 times by default,
+            with a short exponential backoff. Connection errors (for example, due to a network connectivity problem),
+            408 Request Timeout, 409 Conflict, 429 Rate Limit, and >=500 Internal errors are all retried by default.
+
+    Returns:
+        Callable[..., Any]: A callable that creates a client and returns completion results
+
+
+    Raises:
+        InvalidOAIAPITypeError: If an invalid API type string is provided.
+        NotImplementedError: If the specified API type is recognized but not yet supported (e.g., 'azure_openai').
+    """
+    if api_type not in ["openai", "azure_openai"]:
+        raise InvalidOAIAPITypeError(f"Invalid OAIAPIType: {api_type}. Must be 'openai' or 'azure_openai'.")
+
+    if api_type == "openai":
+        return create_client_callable(
+            OpenAI,
+            api_key=api_key,
+            organization=organization,
+            timeout=timeout,
+            max_retries=max_retries,
+        )
+    elif api_type == "azure_openai":
+        if api_key:
+            return create_client_callable(
+                AzureOpenAI,
+                api_version=aoai_api_version,
+                azure_endpoint=azure_endpoint,
+                api_key=api_key,
+                timeout=timeout,
+                max_retries=max_retries,
+            )
+        else:
+            azure_credential = DefaultAzureCredential()
+            ad_token_provider = get_bearer_token_provider(
+                azure_credential, "https://cognitiveservices.azure.com/.default"
+            )
+            return create_client_callable(
+                AzureOpenAI,
+                api_version=aoai_api_version,
+                azure_endpoint=azure_endpoint,
+                azure_ad_token_provider=ad_token_provider,
+                timeout=timeout,
+                max_retries=max_retries,
+            )
+    else:
+        raise NotImplementedError(f"API type '{api_type}' is invalid.")

not_again_ai/llm/embedding/types.py

@@ -0,0 +1,23 @@
+from typing import Any
+
+from pydantic import BaseModel, Field
+
+
+class EmbeddingRequest(BaseModel):
+    input: str | list[str]
+    model: str
+    dimensions: int | None = Field(default=None)
+
+
+class EmbeddingObject(BaseModel):
+    embedding: list[float]
+    index: int
+
+
+class EmbeddingResponse(BaseModel):
+    embeddings: list[EmbeddingObject]
+    total_tokens: int | None = Field(default=None)
+    response_duration: float
+
+    errors: str = Field(default="")
+    extras: Any | None = Field(default=None)

not_again_ai/llm/prompting/{compile_messages.py → compile_prompt.py}

@@ -1,4 +1,5 @@
 import base64
+from collections.abc import Sequence
 from copy import deepcopy
 import mimetypes
 from pathlib import Path
@@ -8,10 +9,27 @@ from liquid import Template
 from openai.lib._pydantic import to_strict_json_schema
 from pydantic import BaseModel
 
-from not_again_ai.llm.chat_completion.types import MessageT, TextContent
+from not_again_ai.llm.chat_completion.types import MessageT
 
 
-def compile_messages(messages: list[MessageT], variables: dict[str, str]) -> list[MessageT]:
+def _apply_templates(value: Any, variables: dict[str, str]) -> Any:
+    """Recursively applies Liquid templating to all string fields within the given value."""
+    if isinstance(value, str):
+        return Template(value).render(**variables)
+    elif isinstance(value, list):
+        return [_apply_templates(item, variables) for item in value]
+    elif isinstance(value, dict):
+        return {key: _apply_templates(val, variables) for key, val in value.items()}
+    elif isinstance(value, BaseModel):
+        # Process each field in the BaseModel by converting it to a dict,
+        # applying templating to its values, and then re-instantiating the model.
+        processed_data = {key: _apply_templates(val, variables) for key, val in value.model_dump().items()}
+        return value.__class__(**processed_data)
+    else:
+        return value
+
+
+def compile_messages(messages: Sequence[MessageT], variables: dict[str, str]) -> Sequence[MessageT]:
     """Compiles messages using Liquid templating and the provided variables.
     Calls Template(content_part).render(**variables) on each text content part.
 
@@ -23,19 +41,28 @@ def compile_messages(messages: list[MessageT], variables: dict[str, str]) -> lis
         The same list of messages with the content parts injected with the variables.
     """
     messages_formatted = deepcopy(messages)
-    for message in messages_formatted:
-        if isinstance(message.content, str):
-            # For simple string content, apply template directly
-            message.content = Template(message.content).render(**variables)
-        elif isinstance(message.content, list):
-            # For UserMessage with content parts
-            for content_part in message.content:
-                if isinstance(content_part, TextContent):
-                    content_part.text = Template(content_part.text).render(**variables)
-                # ImageContent parts are left unchanged
+    messages_formatted = [_apply_templates(message, variables) for message in messages_formatted]
     return messages_formatted
 
 
+def compile_tools(tools: Sequence[dict[str, Any]], variables: dict[str, str]) -> Sequence[dict[str, Any]]:
+    """Compiles a list of tool argument dictionaries using Liquid templating and provided variables.
+
+    Each dictionary in the list is deep copied and processed recursively to substitute any Liquid
+    templates present in its data structure.
+
+    Args:
+        tools: A list of dictionaries representing tool arguments, where values can include Liquid templates.
+        variables: A dictionary of variables to substitute into the Liquid templates.
+
+    Returns:
+        A new list of dictionaries with the Liquid templates replaced by their corresponding variable values.
+    """
+    tools_formatted = deepcopy(tools)
+    tools_formatted = [_apply_templates(tool, variables) for tool in tools_formatted]
+    return tools_formatted
+
+
 def encode_image(image_path: Path) -> str:
     """Encodes an image file at the given Path to base64.
 

{not_again_ai-0.15.0.dist-info → not_again_ai-0.16.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: not-again-ai
-Version: 0.15.0
+Version: 0.16.0
 Summary: Designed to once and for all collect all the little things that come up over and over again in AI projects and put them in one place.
 License: MIT
 Author: DaveCoDev
@@ -25,11 +25,11 @@ Requires-Dist: loguru (>=0.7)
 Requires-Dist: numpy (>=2.2) ; extra == "statistics"
 Requires-Dist: numpy (>=2.2) ; extra == "viz"
 Requires-Dist: ollama (>=0.4) ; extra == "llm"
-Requires-Dist: openai (>=1.60) ; extra == "llm"
+Requires-Dist: openai (>=1) ; extra == "llm"
 Requires-Dist: pandas (>=2.2) ; extra == "viz"
 Requires-Dist: playwright (>=1.49) ; extra == "data"
 Requires-Dist: pydantic (>=2.10)
-Requires-Dist: pytest-playwright (>=0.6) ; extra == "data"
+Requires-Dist: pytest-playwright (>=0.7) ; extra == "data"
 Requires-Dist: python-liquid (>=1.12) ; extra == "llm"
 Requires-Dist: scikit-learn (>=1.6) ; extra == "statistics"
 Requires-Dist: scipy (>=1.15) ; extra == "statistics"

{not_again_ai-0.15.0.dist-info → not_again_ai-0.16.0.dist-info}/RECORD

@@ -11,8 +11,14 @@ not_again_ai/llm/chat_completion/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW
 not_again_ai/llm/chat_completion/providers/ollama_api.py,sha256=iBTMyF8edo8uxxrorNPtShzmCXG7m0RlEBunWLSO4Mo,7999
 not_again_ai/llm/chat_completion/providers/openai_api.py,sha256=S7TZhDIQ_xpp3JakRVcd3Gpw2UjeHCETdA9MfRKUjCU,12294
 not_again_ai/llm/chat_completion/types.py,sha256=q8APUWWzwCKL0Rs_zEFfph9uBcwh5nAT0f0rp4crvk0,4039
+not_again_ai/llm/embedding/__init__.py,sha256=wscUfROukvw0M0vYccfaVTdXV0P-eICAT5mqM0LaHHc,182
+not_again_ai/llm/embedding/interface.py,sha256=Hj3UiktXEeCUeMwpIDtRkwBfKgaJSnJvclLNyjwUAtE,1144
+not_again_ai/llm/embedding/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+not_again_ai/llm/embedding/providers/ollama_api.py,sha256=m-OCis9WAUT2baGsGVPzejlive40eSNyO6tHmPh6joM,3201
+not_again_ai/llm/embedding/providers/openai_api.py,sha256=JFFqbq0O5snIEnr9VESdp5xehikQBPbs7nwyE6acFsY,5441
+not_again_ai/llm/embedding/types.py,sha256=J4FFLx35Aow2kOaafDReeY9cUNqhWMjaAk5gXkX7SVk,506
 not_again_ai/llm/prompting/__init__.py,sha256=7YnHro1yH01FLGnao27WyqQDFjNYf9npE5UxoR9YrUU,84
-not_again_ai/llm/prompting/compile_messages.py,sha256=HmVCQ-0iVg8vFWZyppxUf9m_ae5c8rK1Zx8ySPD1Bg8,3452
+not_again_ai/llm/prompting/compile_prompt.py,sha256=lnbTOoTc7PumyP_GhfHaLZHp3UUpSB7VAeWOilS1wpI,4703
 not_again_ai/llm/prompting/interface.py,sha256=SMKYabmu3zTWbEDukU6aLU_JQ88apeBWWOF_qZ0s3ww,1783
 not_again_ai/llm/prompting/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 not_again_ai/llm/prompting/providers/openai_tiktoken.py,sha256=8YrEiK3ZHyKVGiVsJ_Rd6eVdISIvcub7ooj-HB7Prsc,4536
@@ -26,7 +32,7 @@ not_again_ai/viz/distributions.py,sha256=OyWwJaNI6lMRm_iSrhq-CORLNvXfeuLSgDtVo3u
 not_again_ai/viz/scatterplot.py,sha256=5CUOWeknbBOaZPeX9oPin5sBkRKEwk8qeFH45R-9LlY,2292
 not_again_ai/viz/time_series.py,sha256=pOGZqXp_2nd6nKo-PUQNCtmMh__69jxQ6bQibTGLwZA,5212
 not_again_ai/viz/utils.py,sha256=hN7gwxtBt3U6jQni2K8j5m5pCXpaJDoNzGhBBikEU28,238
-not_again_ai-0.15.0.dist-info/LICENSE,sha256=btjOgNGpp-ux5xOo1Gx1MddxeWtT9sof3s3Nui29QfA,1071
-not_again_ai-0.15.0.dist-info/METADATA,sha256=_vGJUluFVmoYQrNwLGMh5NWtH6aiJ5BG8G8hlZ5TRpE,15038
-not_again_ai-0.15.0.dist-info/WHEEL,sha256=IYZQI976HJqqOpQU6PHkJ8fb3tMNBFjg-Cn-pwAbaFM,88
-not_again_ai-0.15.0.dist-info/RECORD,,
+not_again_ai-0.16.0.dist-info/LICENSE,sha256=btjOgNGpp-ux5xOo1Gx1MddxeWtT9sof3s3Nui29QfA,1071
+not_again_ai-0.16.0.dist-info/METADATA,sha256=kvwxTcEi-elRl-LuHyh2QtFLrpYHd-U6HjyuAkHYvWQ,15035
+not_again_ai-0.16.0.dist-info/WHEEL,sha256=IYZQI976HJqqOpQU6PHkJ8fb3tMNBFjg-Cn-pwAbaFM,88
+not_again_ai-0.16.0.dist-info/RECORD,,