add nvidia nim llm support (#13176)

llama-index-integrations/llms/llama-index-llms-nvidia/.gitignore

@@ -0,0 +1,153 @@
+llama_index/_static
+.DS_Store
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+bin/
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+etc/
+include/
+lib/
+lib64/
+parts/
+sdist/
+share/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+.ruff_cache
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+notebooks/
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+pyvenv.cfg
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# Jetbrains
+.idea
+modules/
+*.swp
+
+# VsCode
+.vscode
+
+# pipenv
+Pipfile
+Pipfile.lock
+
+# pyright
+pyrightconfig.json

llama-index-integrations/llms/llama-index-llms-nvidia/BUILD

@@ -0,0 +1,3 @@
+poetry_requirements(
+    name="poetry",
+)

llama-index-integrations/llms/llama-index-llms-nvidia/Makefile

@@ -0,0 +1,17 @@
+GIT_ROOT ?= $(shell git rev-parse --show-toplevel)
+
+help:	## Show all Makefile targets.
+	@grep -E '^[a-zA-Z_-]+:.*?## .*$$' $(MAKEFILE_LIST) | awk 'BEGIN {FS = ":.*?## "}; {printf "\033[33m%-30s\033[0m %s\n", $$1, $$2}'
+
+format:	## Run code autoformatters (black).
+	pre-commit install
+	git ls-files | xargs pre-commit run black --files
+
+lint:	## Run linters: pre-commit (black, ruff, codespell) and mypy
+	pre-commit install && git ls-files | xargs pre-commit run --show-diff-on-failure --files
+
+test:	## Run tests via pytest.
+	pytest tests
+
+watch-docs:	## Build and watch documentation.
+	sphinx-autobuild docs/ docs/_build/html --open-browser --watch $(GIT_ROOT)/llama_index/

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

@@ -0,0 +1,37 @@
+# NVIDIA's LLM connector
+
+Install the connector,
+
+```shell
+pip install llama-index-llms-nvidia
+```
+
+With this connector, you'll be able to connect to and generate from compatible models available as hosted [NVIDIA NIMs](https://ai.nvidia.com), such as:
+
+- Google's [gemma-7b](https://build.nvidia.com/google/gemma-7b)
+- Mistal AI's [mistral-7b-instruct-v0.2](https://build.nvidia.com/mistralai/mistral-7b-instruct-v2)
+- And more!
+
+_First_, get a free API key. Go to https://build.nvidia.com, select a model, click "Get API Key".
+Store this key in your environment as `NVIDIA_API_KEY`.
+
+_Then_, try it out.
+
+```python
+from llama_index.llms.nvidia import NVIDIA
+from llama_index.core.llms import ChatMessage, MessageRole
+
+llm = NVIDIA()
+
+messages = [
+    ChatMessage(
+        role=MessageRole.SYSTEM, content=("You are a helpful assistant.")
+    ),
+    ChatMessage(
+        role=MessageRole.USER,
+        content=("What are the most popular house pets in North America?"),
+    ),
+]
+
+llm.chat(messages)
+```

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

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

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

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

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

@@ -0,0 +1,107 @@
+from typing import (
+    Any,
+    Optional,
+    List,
+    Literal,
+)
+
+from llama_index.core.bridge.pydantic import PrivateAttr, BaseModel
+from llama_index.core.base.llms.generic_utils import (
+    get_from_param_or_env,
+)
+
+from llama_index.llms.nvidia.utils import API_CATALOG_MODELS
+
+from llama_index.llms.openai_like import OpenAILike
+
+DEFAULT_MODEL = "meta/llama3-8b-instruct"
+BASE_URL = "https://integrate.api.nvidia.com/v1/"
+
+
+class Model(BaseModel):
+    id: str
+
+
+class NVIDIA(OpenAILike):
+    """NVIDIA's API Catalog Connector."""
+
+    _mode: str = PrivateAttr("nvidia")
+
+    def __init__(
+        self,
+        model: str = DEFAULT_MODEL,
+        nvidia_api_key: Optional[str] = None,
+        api_key: Optional[str] = None,
+        **kwargs: Any,
+    ) -> None:
+        api_key = get_from_param_or_env(
+            "api_key",
+            nvidia_api_key or api_key,
+            "NVIDIA_API_KEY",
+            "NO_API_KEY_PROVIDED",
+        )
+
+        super().__init__(
+            model=model,
+            api_key=api_key,
+            api_base=BASE_URL,
+            is_chat_model=True,
+            default_headers={"User-Agent": "llama-index-llms-nvidia"},
+            **kwargs,
+        )
+
+    @property
+    def available_models(self) -> List[Model]:
+        ids = API_CATALOG_MODELS.keys()
+        if self._mode == "nim":
+            ids = [model.id for model in self._get_client().models.list()]
+        return [Model(id=name) for name in ids]
+
+    @classmethod
+    def class_name(cls) -> str:
+        return "NVIDIA"
+
+    def mode(
+        self,
+        mode: Optional[Literal["nvidia", "nim"]] = "nvidia",
+        *,
+        base_url: Optional[str] = None,
+        model: Optional[str] = None,
+        api_key: Optional[str] = None,
+    ) -> "NVIDIA":
+        """
+        Change the mode.
+
+        There are two modes, "nvidia" and "nim". The "nvidia" mode is the default
+        mode and is used to interact with hosted NIMs. The "nim" mode is used to
+        interact with NVIDIA NIM endpoints, which are typically hosted on-premises.
+
+        For the "nvidia" mode, the "api_key" parameter is available to specify
+        your API key. If not specified, the NVIDIA_API_KEY environment variable
+        will be used.
+
+        For the "nim" mode, the "base_url" parameter is required and the "model"
+        parameter may be necessary. Set base_url to the url of your local NIM
+        endpoint. For instance, "https://localhost:9999/v1". Additionally, the
+        "model" parameter must be set to the name of the model inside the NIM.
+        """
+        if mode == "nim":
+            if not base_url:
+                raise ValueError("base_url is required for nim mode")
+        if mode == "nvidia":
+            api_key = get_from_param_or_env(
+                "api_key",
+                api_key,
+                "NVIDIA_API_KEY",
+            )
+            base_url = base_url or BASE_URL
+
+        self._mode = mode
+        if base_url:
+            self.api_base = base_url
+        if model:
+            self.model = model
+        if api_key:
+            self.api_key = api_key
+
+        return self

llama-index-integrations/llms/llama-index-llms-nvidia/llama_index/llms/nvidia/utils.py

@@ -0,0 +1,19 @@
+from typing import Dict, Optional
+
+API_CATALOG_MODELS: Dict[str, int] = {
+    "mistralai/mistral-7b-instruct-v0.2": 16384,
+    "mistralai/mixtral-8x7b-instruct-v0.1": 16384,
+    "mistralai/mixtral-8x22b-instruct-v0.1": 32768,
+    "mistralai/mistral-large": 16384,
+    "google/gemma-7b": 4096,
+    "google/gemma-2b": 4096,
+    "google/codegemma-7b": 4096,
+    "meta/llama2-70b": 1024,
+    "meta/codellama-70b": 1024,
+    "meta/llama3-8b-instruct": 6000,
+    "meta/llama3-70b-instruct": 6000,
+}
+
+
+def catalog_modelname_to_contextsize(modelname: str) -> Optional[int]:
+    return API_CATALOG_MODELS.get(modelname, None)