Contextual Generate model

Signed-off-by: Sean Smith <[email protected]>

llama-index-integrations/llms/llama-index-llms-contextual/README.md

@@ -0,0 +1,19 @@
+# Contextual LLM Integration for LlamaIndex
+
+This package provides a Contextual LLM integration for LlamaIndex.
+
+## Installation
+
+```bash
+pip install llama-index-llms-contextual
+```
+
+## Usage
+
+```python
+from llama_index.llms.contextual import Contextual
+
+llm = Contextual(model="contextual-clm", api_key="your_api_key")
+
+response = llm.complete("Explain the importance of Grounded Language Models.")
+```

llama-index-integrations/llms/llama-index-llms-contextual/llama_index/llms/contextual/BUILD

@@ -0,0 +1 @@
+python_sources()

llama-index-integrations/llms/llama-index-llms-contextual/llama_index/llms/contextual/__init__.py

@@ -0,0 +1,3 @@
+from llama_index.llms.contextual.base import Contextual
+
+__all__ = ["Contextual"]

llama-index-integrations/llms/llama-index-llms-contextual/llama_index/llms/contextual/base.py

@@ -0,0 +1,117 @@
+import os
+from typing import Any, Optional
+
+from llama_index.llms.openai_like import OpenAILike
+from pydantic import Field
+from llama_index.core.llms.callbacks import (
+    llm_chat_callback,
+    llm_completion_callback,
+)
+from llama_index.core.base.llms.types import (
+    CompletionResponse,
+    CompletionResponseGen,
+    ChatResponse,
+    ChatResponseGen,
+    ChatResponseAsyncGen,
+    CompletionResponseAsyncGen,
+    LLMMetadata,
+    MessageRole,
+    ChatMessage,
+)
+
+
+from contextual import ContextualAI
+
+class Contextual(OpenAILike):
+    """
+    Generate a response using Contextual's Grounded Language Model (GLM), an LLM engineered specifically to prioritize faithfulness to in-context retrievals over parametric knowledge to reduce hallucinations in Retrieval-Augmented Generation.
+
+    The total request cannot exceed 32,000 tokens. Email [email protected] with any feedback or questions.
+
+    Examples:
+        `pip install llama-index-llms-contextual`
+
+        ```python
+        from llama_index.llms.contextual import Contextual
+
+        # Set up the Contextual class with the required model and API key
+        llm = Contextual(model="contextual-clm", api_key="your_api_key")
+
+        # Call the complete method with a query
+        response = llm.complete("Explain the importance of low latency LLMs")
+
+        print(response)
+        ```
+    """
+
+    model: str = Field(description="The model to use. Currently only supports `v1`.")
+    api_key: str = Field(description="The API key to use.", default=os.environ.get("API_KEY", None))
+    base_url: str = Field(description="The base URL to use.", default="https://api.contextual.com")
+    system_prompt: str = Field(description="Instructions that the model follows when generating responses. Note that we do not guarantee that the model follows these instructions exactly.")
+    avoid_commentary: bool = Field(description="Flag to indicate whether the model should avoid providing additional commentary in responses. Commentary is conversational in nature and does not contain verifiable claims; therefore, commentary is not strictly grounded in available context. However, commentary may provide useful context which improves the helpfulness of responses.", default=False)
+    client: Any = Field(default=None, exclude=True, description="Contextual AI Client")
+
+    def __init__(
+        self,
+        model: str,
+        api_key: Optional[str] = None,
+        base_url: str = "https://api.contextual.com",
+        system_prompt: str = "",
+        avoid_commentary: bool = False,
+        **openai_llm_kwargs: Any,
+    ) -> None:
+        api_key = api_key or os.environ.get("API_KEY", None)
+
+        try:
+            self.client = ContextualAI(api_key=api_key, base_url=base_url)
+        except Exception as e:
+            raise ValueError(f"Failed to initialize Contextual client: {e}")
+
+
+        super().__init__(
+            model=model,
+            api_key=api_key,
+            api_base=base_url,
+            is_chat_model=openai_llm_kwargs.pop("is_chat_model", True),
+            **openai_llm_kwargs,
+        )
+
+    @classmethod
+    def class_name(cls) -> str:
+        """Get class name."""
+        return "contextual-clm"
+
+    # Synchronous Methods
+    @llm_completion_callback()
+    def complete(self, prompt: str, **kwargs) -> CompletionResponse:
+        """
+        Generate completion for the given prompt.
+
+        Args:
+            prompt (str): The input prompt to generate completion for.
+            **kwargs: Additional keyword arguments for the API request.
+
+        Returns:
+            str: The generated text completion.
+        """
+        return self._generate(
+            knowledge=None,
+            messages=[ChatMessage(role=MessageRole.USER, content=prompt)],
+            model=self.model,
+            system_prompt=self.system_prompt,
+            **kwargs,
+        )
+
+    def _generate(self, knowledge, messages, model, system_prompt, **kwargs) -> CompletionResponse:
+        """
+        Generate completion for the given prompt.
+        """
+        raw_message = self.client.generate.create(
+            messages=messages,
+            knowledge=knowledge,
+            model=self.model_name,
+            system_prompt=system_prompt,
+            extra_body={
+                "avoid_commentary": self.avoid_commentary,
+            },
+        )

llama-index-integrations/llms/llama-index-llms-contextual/llama_index/llms/contextual/test-contextual.ipynb

@@ -0,0 +1,52 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Contextual LLM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "vscode": {
+     "languageId": "plaintext"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "!pip install llama-index-llms-deepseek"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {
+    "vscode": {
+     "languageId": "plaintext"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "from llama_index.llms.contextual import Contextual\n",
+    "\n",
+    "# Set up the Contextual class with the required model and API key\n",
+    "llm = Contextual(model=\"v1\", api_key=\"your_api_key\")\n",
+    "\n",
+    "# Call the complete method with a query\n",
+    "response = llm.complete(\"Explain the importance of Grounded Language Models.\")\n",
+    "\n",
+    "print(response)"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

llama-index-integrations/llms/llama-index-llms-contextual/pyproject.toml

@@ -0,0 +1,46 @@
+[build-system]
+build-backend = "poetry.core.masonry.api"
+requires = ["poetry-core"]
+
+[tool.codespell]
+check-filenames = true
+check-hidden = true
+skip = "*.csv,*.html,*.json,*.jsonl,*.pdf,*.txt,*.ipynb"
+
+[tool.llamahub]
+contains_example = false
+import_path = "llama_index.llms.contextual"
+
+[tool.llamahub.class_authors]
+Contextual = "sean-smith"
+
+[tool.mypy]
+disallow_untyped_defs = true
+exclude = ["_static", "build", "examples", "notebooks", "venv"]
+ignore_missing_imports = true
+python_version = "3.8"
+
+[tool.poetry]
+authors = ["Sean Smith <[email protected]>"]
+description = "llama-index contextual integration"
+exclude = ["**/BUILD"]
+license = "MIT"
+name = "llama-index-llms-contextual"
+readme = "README.md"
+version = "0.0.1"
+
+[tool.poetry.dependencies]
+python = ">=3.9,<4.0"
+llama-index-llms-openai-like = "^0.3.3"
+contextual-client = "^0.2.0"
+
+[tool.poetry.group.dev.dependencies.black]
+extras = ["jupyter"]
+version = "<=23.9.1,>=23.7.0"
+
+[tool.poetry.group.dev.dependencies.codespell]
+extras = ["toml"]
+version = ">=v2.2.6"
+
+[[tool.poetry.packages]]
+include = "llama_index/"

llama-index-integrations/llms/llama-index-llms-contextual/tests/BUILD

@@ -0,0 +1 @@
+python_sources()

llama-index-integrations/llms/llama-index-llms-contextual/tests/test.py

@@ -0,0 +1,9 @@
+from llama_index.llms.contextual import Contextual
+
+# Set up the Contextual class with the required model and API key
+llm = Contextual(model="contextual-clm", api_key="key-47EdcA3Tim9bif-W96XAsA0nFPf9ZPvkwtE4cL7vP5s7NtAHw")
+
+# Call the complete method with a query
+response = llm.complete("Explain the importance of low latency LLMs")
+
+print(response)