Docs for meta-reference-gpu

distributions/meta-reference-gpu/run-with-safety.yaml

@@ -0,0 +1,70 @@
+version: '2'
+image_name: meta-reference-gpu
+docker_image: null
+conda_env: null
+apis:
+- agents
+- inference
+- memory
+- safety
+- telemetry
+providers:
+  inference:
+  - provider_id: meta-reference-inference
+    provider_type: inline::meta-reference
+    config:
+      model: ${env.INFERENCE_MODEL}
+      max_seq_len: 4096
+      checkpoint_dir: ${env.INFERENCE_CHECKPOINT_DIR:null}
+  - provider_id: meta-reference-safety
+    provider_type: inline::meta-reference
+    config:
+      model: ${env.SAFETY_MODEL}
+      max_seq_len: 4096
+      checkpoint_dir: ${env.SAFETY_CHECKPOINT_DIR:null}
+  memory:
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/meta-reference-gpu}/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/meta-reference-gpu}/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/meta-reference-gpu}/registry.db
+models:
+- metadata: {}
+  model_id: ${env.INFERENCE_MODEL}
+  provider_id: meta-reference-inference
+  provider_model_id: null
+- metadata: {}
+  model_id: ${env.SAFETY_MODEL}
+  provider_id: meta-reference-safety
+  provider_model_id: null
+shields:
+- params: null
+  shield_id: ${env.SAFETY_MODEL}
+  provider_id: null
+  provider_shield_id: null
+memory_banks: []
+datasets: []
+scoring_fns: []
+eval_tasks: []

distributions/meta-reference-gpu/run.yaml

@@ -1,68 +1,56 @@
 version: '2'
-image_name: local
+image_name: meta-reference-gpu
 docker_image: null
-conda_env: local
+conda_env: null
 apis:
-- shields
 - agents
-- models
-- memory
-- memory_banks
 - inference
+- memory
 - safety
+- telemetry
 providers:
   inference:
-  - provider_id: inference0
+  - provider_id: meta-reference-inference
     provider_type: inline::meta-reference
     config:
-      model: Llama3.2-3B-Instruct
-      quantization: null
-      torch_seed: null
+      model: ${env.INFERENCE_MODEL}
       max_seq_len: 4096
-      max_batch_size: 1
-  - provider_id: inference1
-    provider_type: inline::meta-reference
-    config:
-      model: Llama-Guard-3-1B
-      quantization: null
-      torch_seed: null
-      max_seq_len: 2048
-      max_batch_size: 1
-  safety:
-  - provider_id: meta0
-    provider_type: inline::llama-guard
-    config:
-      model: Llama-Guard-3-1B
-      excluded_categories: []
-  - provider_id: meta1
-    provider_type: inline::prompt-guard
-    config:
-      model: Prompt-Guard-86M
-# Uncomment to use prompt guard
-#      prompt_guard_shield:
-#        model: Prompt-Guard-86M
+      checkpoint_dir: ${env.INFERENCE_CHECKPOINT_DIR:null}
   memory:
-  - provider_id: meta0
-    provider_type: inline::meta-reference
+  - provider_id: faiss
+    provider_type: inline::faiss
+    config:
+      kvstore:
+        type: sqlite
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/meta-reference-gpu}/faiss_store.db
+  safety:
+  - provider_id: llama-guard
+    provider_type: inline::llama-guard
     config: {}
-  # Uncomment to use pgvector
-  # - provider_id: pgvector
-  #   provider_type: remote::pgvector
-  #   config:
-  #     host: 127.0.0.1
-  #     port: 5432
-  #     db: postgres
-  #     user: postgres
-  #     password: mysecretpassword
   agents:
-  - provider_id: meta0
+  - provider_id: meta-reference
     provider_type: inline::meta-reference
     config:
       persistence_store:
-        namespace: null
         type: sqlite
-        db_path: ~/.llama/runtime/agents_store.db
+        namespace: null
+        db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/meta-reference-gpu}/agents_store.db
   telemetry:
-  - provider_id: meta0
+  - provider_id: meta-reference
     provider_type: inline::meta-reference
     config: {}
+metadata_store:
+  namespace: null
+  type: sqlite
+  db_path: ${env.SQLITE_STORE_DIR:~/.llama/distributions/meta-reference-gpu}/registry.db
+models:
+- metadata: {}
+  model_id: ${env.INFERENCE_MODEL}
+  provider_id: meta-reference-inference
+  provider_model_id: null
+shields: []
+memory_banks: []
+datasets: []
+scoring_fns: []
+eval_tasks: []

docs/source/getting_started/distributions/self_hosted_distro/meta-reference-gpu.md

@@ -1,15 +1,23 @@
 # Meta Reference Distribution
 
-The `llamastack/distribution-meta-reference-gpu` distribution consists of the following provider configurations.
+The `llamastack/distribution-meta-reference-gpu` distribution consists of the following provider configurations:
+
+| API | Provider(s) |
+|-----|-------------|
+| agents | `inline::meta-reference` |
+| inference | `inline::meta-reference` |
+| memory | `inline::faiss`, `remote::chromadb`, `remote::pgvector` |
+| safety | `inline::llama-guard` |
+| telemetry | `inline::meta-reference` |
 
 
-| **API**         	| **Inference** 	| **Agents**     	| **Memory**                                       	| **Safety**     	| **Telemetry**  	|
-|-----------------	|---------------	|----------------	|--------------------------------------------------	|----------------	|----------------	|
-| **Provider(s)** 	| meta-reference  	| meta-reference 	| meta-reference, remote::pgvector, remote::chroma 	| meta-reference 	| meta-reference 	|
+Note that you need access to nvidia GPUs to run this distribution. This distribution is not compatible with CPU-only machines or machines with AMD GPUs.
 
 
-### Step 0. Prerequisite - Downloading Models
-Please make sure you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/cli_reference/download_models.html) here to download the models.
+
+## Prerequisite: Downloading Models
+
+Please make sure you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/cli_reference/download_models.html) here to download the models. Run `llama model list` to see the available models to download, and `llama model download` to download the checkpoints.
 
 ```
 $ ls ~/.llama/checkpoints
@@ -17,55 +25,56 @@ Llama3.1-8B           Llama3.2-11B-Vision-Instruct  Llama3.2-1B-Instruct  Llama3
 Llama3.1-8B-Instruct  Llama3.2-1B                   Llama3.2-3B-Instruct  Llama-Guard-3-1B              Prompt-Guard-86M
 ```
 
-### Step 1. Start the Distribution
+## Running the Distribution
 
-#### (Option 1) Start with Docker
-```
-$ cd distributions/meta-reference-gpu && docker compose up
+You can do this via Conda (build code) or Docker which has a pre-built image.
+
+### Via Docker
+
+This method allows you to get started quickly without having to build the distribution code.
+
+```bash
+LLAMA_STACK_PORT=5001
+docker run \
+  -it \
+  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
+  -v ./run.yaml:/root/my-run.yaml \
+  llamastack/distribution-meta-reference-gpu \
+  /root/my-run.yaml \
+  --port $LLAMA_STACK_PORT \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct
 ```
 
-> [!NOTE]
-> This assumes you have access to GPU to start a local server with access to your GPU.
+If you are using Llama Stack Safety / Shield APIs, use:
 
-
-> [!NOTE]
-> `~/.llama` should be the path containing downloaded weights of Llama models.
-
-
-This will download and start running a pre-built docker container. Alternatively, you may use the following commands:
-
-```
-docker run -it -p 5000:5000 -v ~/.llama:/root/.llama -v ./run.yaml:/root/my-run.yaml --gpus=all distribution-meta-reference-gpu --yaml_config /root/my-run.yaml
+```bash
+docker run \
+  -it \
+  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
+  -v ./run-with-safety.yaml:/root/my-run.yaml \
+  llamastack/distribution-meta-reference-gpu \
+  /root/my-run.yaml \
+  --port $LLAMA_STACK_PORT \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct \
+  --env SAFETY_MODEL=meta-llama/Llama-Guard-3-1B
 ```
 
-#### (Option 2) Start with Conda
+### Via Conda
 
-1. Install the `llama` CLI. See [CLI Reference](https://llama-stack.readthedocs.io/en/latest/cli_reference/index.html)
+Make sure you have done `pip install llama-stack` and have the Llama Stack CLI available.
 
-2. Build the `meta-reference-gpu` distribution
-
-```
-$ llama stack build --template meta-reference-gpu --image-type conda
+```bash
+llama stack build --template meta-reference-gpu --image-type conda
+llama stack run ./run.yaml \
+  --port 5001 \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct
 ```
 
-3. Start running distribution
-```
-$ cd distributions/meta-reference-gpu
-$ llama stack run ./run.yaml
-```
+If you are using Llama Stack Safety / Shield APIs, use:
 
-### (Optional) Serving a new model
-You may change the `config.model` in `run.yaml` to update the model currently being served by the distribution. Make sure you have the model checkpoint downloaded in your `~/.llama`.
+```bash
+llama stack run ./run-with-safety.yaml \
+  --port 5001 \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct \
+  --env SAFETY_MODEL=meta-llama/Llama-Guard-3-1B
 ```
-inference:
-  - provider_id: meta0
-    provider_type: inline::meta-reference
-    config:
-      model: Llama3.2-11B-Vision-Instruct
-      quantization: null
-      torch_seed: null
-      max_seq_len: 4096
-      max_batch_size: 1
-```
-
-Run `llama model list` to see the available models to download, and `llama model download` to download the checkpoints.

llama_stack/providers/inline/inference/meta_reference/config.py

@@ -49,6 +49,18 @@ class MetaReferenceInferenceConfig(BaseModel):
         resolved = resolve_model(self.model)
         return resolved.pth_file_count
 
+    @classmethod
+    def sample_run_config(
+        cls,
+        model: str = "Llama3.2-3B-Instruct",
+        checkpoint_dir: str = "${env.CHECKPOINT_DIR:null}",
+    ) -> Dict[str, Any]:
+        return {
+            "model": model,
+            "max_seq_len": 4096,
+            "checkpoint_dir": checkpoint_dir,
+        }
+
 
 class MetaReferenceQuantizedInferenceConfig(MetaReferenceInferenceConfig):
     quantization: QuantizationConfig

llama_stack/providers/inline/inference/meta_reference/generation.py

@@ -107,7 +107,7 @@ class Llama:
             sys.stdout = open(os.devnull, "w")
 
         start_time = time.time()
-        if config.checkpoint_dir:
+        if config.checkpoint_dir and config.checkpoint_dir != "null":
             ckpt_dir = config.checkpoint_dir
         else:
             ckpt_dir = model_checkpoint_dir(model)
@@ -137,7 +137,6 @@ class Llama:
         ), f"model_args vocab = {model_args.vocab_size} but tokenizer vocab = {tokenizer.n_words}"
 
         if isinstance(config, MetaReferenceQuantizedInferenceConfig):
-
             if isinstance(config.quantization, Fp8QuantizationConfig):
                 from .quantization.loader import convert_to_fp8_quantized_model
 

llama_stack/templates/meta-reference-gpu/__init__.py

@@ -0,0 +1,7 @@
+# Copyright (c) Meta Platforms, 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 .meta_reference import get_distribution_template  # noqa: F401

llama_stack/templates/meta-reference-gpu/build.yaml

@@ -1,13 +1,19 @@
+version: '2'
 name: meta-reference-gpu
 distribution_spec:
-  docker_image: pytorch/pytorch:2.5.0-cuda12.4-cudnn9-runtime
-  description: Use code from `llama_stack` itself to serve all llama stack APIs
+  description: Use Meta Reference for running LLM inference
+  docker_image: null
   providers:
-    inference: meta-reference
+    inference:
+    - inline::meta-reference
     memory:
     - inline::faiss
     - remote::chromadb
     - remote::pgvector
-    safety: inline::llama-guard
-    agents: inline::meta-reference
-    telemetry: inline::meta-reference
+    safety:
+    - inline::llama-guard
+    agents:
+    - inline::meta-reference
+    telemetry:
+    - inline::meta-reference
+image_type: conda

llama_stack/templates/meta-reference-gpu/doc_template.md

@@ -0,0 +1,82 @@
+# Meta Reference Distribution
+
+The `llamastack/distribution-{{ name }}` distribution consists of the following provider configurations:
+
+{{ providers_table }}
+
+Note that you need access to nvidia GPUs to run this distribution. This distribution is not compatible with CPU-only machines or machines with AMD GPUs.
+
+{% if run_config_env_vars %}
+### Environment Variables
+
+The following environment variables can be configured:
+
+{% for var, (default_value, description) in run_config_env_vars.items() %}
+- `{{ var }}`: {{ description }} (default: `{{ default_value }}`)
+{% endfor %}
+{% endif %}
+
+
+## Prerequisite: Downloading Models
+
+Please make sure you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/cli_reference/download_models.html) here to download the models. Run `llama model list` to see the available models to download, and `llama model download` to download the checkpoints.
+
+```
+$ ls ~/.llama/checkpoints
+Llama3.1-8B           Llama3.2-11B-Vision-Instruct  Llama3.2-1B-Instruct  Llama3.2-90B-Vision-Instruct  Llama-Guard-3-8B
+Llama3.1-8B-Instruct  Llama3.2-1B                   Llama3.2-3B-Instruct  Llama-Guard-3-1B              Prompt-Guard-86M
+```
+
+## Running the Distribution
+
+You can do this via Conda (build code) or Docker which has a pre-built image.
+
+### Via Docker
+
+This method allows you to get started quickly without having to build the distribution code.
+
+```bash
+LLAMA_STACK_PORT=5001
+docker run \
+  -it \
+  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
+  -v ./run.yaml:/root/my-run.yaml \
+  llamastack/distribution-{{ name }} \
+  /root/my-run.yaml \
+  --port $LLAMA_STACK_PORT \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct
+```
+
+If you are using Llama Stack Safety / Shield APIs, use:
+
+```bash
+docker run \
+  -it \
+  -p $LLAMA_STACK_PORT:$LLAMA_STACK_PORT \
+  -v ./run-with-safety.yaml:/root/my-run.yaml \
+  llamastack/distribution-{{ name }} \
+  /root/my-run.yaml \
+  --port $LLAMA_STACK_PORT \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct \
+  --env SAFETY_MODEL=meta-llama/Llama-Guard-3-1B
+```
+
+### Via Conda
+
+Make sure you have done `pip install llama-stack` and have the Llama Stack CLI available.
+
+```bash
+llama stack build --template meta-reference-gpu --image-type conda
+llama stack run ./run.yaml \
+  --port 5001 \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct
+```
+
+If you are using Llama Stack Safety / Shield APIs, use:
+
+```bash
+llama stack run ./run-with-safety.yaml \
+  --port 5001 \
+  --env INFERENCE_MODEL=meta-llama/Llama-3.2-3B-Instruct \
+  --env SAFETY_MODEL=meta-llama/Llama-Guard-3-1B
+```

llama_stack/templates/meta-reference-gpu/meta_reference.py

@@ -0,0 +1,100 @@
+# Copyright (c) Meta Platforms, 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_stack.distribution.datatypes import ModelInput, Provider, ShieldInput
+from llama_stack.providers.inline.inference.meta_reference import (
+    MetaReferenceInferenceConfig,
+)
+from llama_stack.templates.template import DistributionTemplate, RunConfigSettings
+
+
+def get_distribution_template() -> DistributionTemplate:
+    providers = {
+        "inference": ["inline::meta-reference"],
+        "memory": ["inline::faiss", "remote::chromadb", "remote::pgvector"],
+        "safety": ["inline::llama-guard"],
+        "agents": ["inline::meta-reference"],
+        "telemetry": ["inline::meta-reference"],
+    }
+
+    inference_provider = Provider(
+        provider_id="meta-reference-inference",
+        provider_type="inline::meta-reference",
+        config=MetaReferenceInferenceConfig.sample_run_config(
+            model="${env.INFERENCE_MODEL}",
+            checkpoint_dir="${env.INFERENCE_CHECKPOINT_DIR:null}",
+        ),
+    )
+
+    inference_model = ModelInput(
+        model_id="${env.INFERENCE_MODEL}",
+        provider_id="meta-reference-inference",
+    )
+    safety_model = ModelInput(
+        model_id="${env.SAFETY_MODEL}",
+        provider_id="meta-reference-safety",
+    )
+
+    return DistributionTemplate(
+        name="meta-reference-gpu",
+        distro_type="self_hosted",
+        description="Use Meta Reference for running LLM inference",
+        template_path=Path(__file__).parent / "doc_template.md",
+        providers=providers,
+        default_models=[inference_model, safety_model],
+        run_configs={
+            "run.yaml": RunConfigSettings(
+                provider_overrides={
+                    "inference": [inference_provider],
+                },
+                default_models=[inference_model],
+            ),
+            "run-with-safety.yaml": RunConfigSettings(
+                provider_overrides={
+                    "inference": [
+                        inference_provider,
+                        Provider(
+                            provider_id="meta-reference-safety",
+                            provider_type="inline::meta-reference",
+                            config=MetaReferenceInferenceConfig.sample_run_config(
+                                model="${env.SAFETY_MODEL}",
+                                checkpoint_dir="${env.SAFETY_CHECKPOINT_DIR:null}",
+                            ),
+                        ),
+                    ],
+                },
+                default_models=[
+                    inference_model,
+                    safety_model,
+                ],
+                default_shields=[ShieldInput(shield_id="${env.SAFETY_MODEL}")],
+            ),
+        },
+        docker_compose_env_vars={
+            "LLAMASTACK_PORT": (
+                "5001",
+                "Port for the Llama Stack distribution server",
+            ),
+            "INFERENCE_MODEL": (
+                "meta-llama/Llama-3.2-3B-Instruct",
+                "Inference model loaded into the Meta Reference server",
+            ),
+            "INFERENCE_CHECKPOINT_DIR": (
+                "null",
+                "Directory containing the Meta Reference model checkpoint",
+            ),
+            "SAFETY_MODEL": (
+                "meta-llama/Llama-Guard-3-1B",
+                "Name of the safety (Llama-Guard) model to use",
+            ),
+            "SAFETY_CHECKPOINT_DIR": (
+                "null",
+                "Directory containing the Llama-Guard model checkpoint",
+            ),
+        },
+    )