Nutanix AI distribution

llama_stack/templates/nutanix/__init__.py

@@ -0,0 +1,7 @@
+# Copyright (c) Nutanix, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from .nutanix import get_distribution_template  # noqa: F401

llama_stack/templates/nutanix/build.yaml

@@ -3,7 +3,10 @@ distribution_spec:
   description: Use Nutanix AI Endpoint for running LLM inference
   providers:
     inference: remote::nutanix
-    memory: inline::faiss
+    memory:
+      - inline::faiss
+      - remote::chromadb
+      - remote::pgvector
     safety: inline::llama-guard
     agents: inline::meta-reference
     telemetry: inline::meta-reference

llama_stack/templates/nutanix/doc_template.md

@@ -0,0 +1,40 @@
+# Nutanix Distribution
+
+The `llamastack/distribution-nutanix` distribution consists of the following provider configurations.
+
+
+| **API**         	| **Inference** 	| **Agents**     	| **Memory**                                       	| **Safety**     	| **Telemetry**  	|
+|-----------------	|---------------	|----------------	|--------------------------------------------------	|----------------	|----------------	|
+| **Provider(s)** 	| remote::nutanix   	| meta-reference 	| meta-reference 	| meta-reference 	| meta-reference 	|
+
+
+### Start the Distribution (Hosted remote)
+
+> [!NOTE]
+> This assumes you have an hosted Nutanix AI endpoint and an API Key.
+
+1. Clone the repo
+```
+git clone git@github.com:meta-llama/llama-stack.git
+cd llama-stack
+```
+
+2. Config the model name
+
+Please adjust the `NUTANIX_SUPPORTED_MODELS` variable at line 29 in `llama_stack/providers/adapters/inference/nutanix/nutanix.py` according to your deployment.
+
+3. Build the distrbution
+```
+pip install -e .
+llama stack build --template nutanix --image-type conda
+```
+
+4. Edit the yaml file
+```
+vim
+```
+
+5. Serve and enjoy!
+```
+llama stack run ntnx --port 174
+```

llama_stack/templates/nutanix/nutanix.py

@@ -0,0 +1,70 @@
+# Copyright (c) Nutanix, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the terms described in the LICENSE file in
+# the root directory of this source tree.
+
+from pathlib import Path
+
+from llama_models.sku_list import all_registered_models
+
+from llama_stack.distribution.datatypes import ModelInput, Provider, ShieldInput
+from llama_stack.templates.template import DistributionTemplate, RunConfigSettings
+
+from llama_stack.providers.remote.inference.nutanix import NutanixImplConfig
+from llama_stack.providers.remote.inference.nutanix.nutanix import MODEL_ALIASES
+
+def get_distribution_template() -> DistributionTemplate:
+    providers = {
+        "inference": ["remote::nutanix"],
+        "memory": ["inline::faiss", "remote::chromadb", "remote::pgvector"],
+        "safety": ["inline::llama-guard"],
+        "agents": ["inline::meta-reference"],
+        "telemetry": ["inline::meta-reference"],
+    }
+    
+    inference_provider = Provider(
+        provider_id="nutanix",
+        provider_type="remote::nutanix",
+        config=NutanixImplConfig.sample_run_config(),
+    )
+    
+    core_model_to_hf_repo = {
+        m.descriptor(): m.huggingface_repo for m in all_registered_models()
+    }
+    default_models = [
+        ModelInput(
+            model_id=core_model_to_hf_repo[m.llama_model],
+            provider_model_id=m.provider_model_id,
+        )
+        for m in MODEL_ALIASES
+    ]
+
+    return DistributionTemplate(
+        name="nutanix",
+        distro_type="remote::nutanix",
+        description="Use Nutanix AI Endpoint for running LLM inference",
+        docker_image=None,
+        template_path=Path(__file__).parent / "doc_template.md",
+        providers=providers,
+        default_models=default_models,
+        run_configs={
+            "run.yaml": RunConfigSettings(
+                provider_overrides={
+                    "inference": [inference_provider],
+                },
+                default_models=default_models,
+                default_shields=[ShieldInput(shield_id="meta-llama/Llama-Guard-3-8B")],
+            ),
+        },
+        run_config_env_vars={
+            "LLAMASTACK_PORT": (
+                "5001",
+                "Port for the Llama Stack distribution server",
+            ),
+            "NUTANIX_API_KEY": (
+                "",
+                "Nutanix AI Endpoint API Key",
+            ),
+        },
+    )

llama_stack/templates/nutanix/run.yaml

@@ -0,0 +1,54 @@
+version: '2'
+image_name: nutanix
+docker_image: null
+conda_env: nutanix
+apis:
+- agents
+- inference
+- memory
+- safety
+- telemetry
+providers:
+  inference:
+  - provider_id: nutanix
+    provider_type: remote::nutanix
+    config: 
+      url: https://ai.nutanix.com/api/v1
+      api_key: ${env.NUTANIX_API_KEY}
+  memory:
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nutanix}/faiss_store.db
+  safety:
+  - provider_id: llama-guard
+    provider_type: inline::llama-guard
+    config: {}
+  agents:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config:
+      persistence_store:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nutanix}/agents_store.db
+  telemetry:
+  - provider_id: meta-reference
+    provider_type: inline::meta-reference
+    config: {}
+metadata_store:
+  namespace: null
+  type: sqlite
+  db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/nutanix}/registry.db
+models:
+  - model_id: meta-llama/Llama-3.1-8B-Instruct
+    provider_id: null
+    provider_model_id: vllm-llama-3-1
+shields: []
+memory_banks: []
+datasets: []
+scoring_fns: []
+eval_tasks: []