Merge branch 'main' into pr1334

docs/openapi_generator/generate.py

@@ -55,6 +55,7 @@ def main(output_dir: str):
                 a set of endpoints and their corresponding interfaces that are tailored to
                 best leverage Llama Models.""",
             ),
+            include_standard_error_responses=True,
         ),
     )
 

docs/openapi_generator/pyopenapi/generator.py

@@ -10,6 +10,7 @@ import typing
 from dataclasses import make_dataclass
 from typing import Any, Dict, Set, Union
 
+from llama_stack.apis.datatypes import Error
 from llama_stack.strong_typing.core import JsonType
 from llama_stack.strong_typing.docstring import Docstring, parse_type
 from llama_stack.strong_typing.inspection import (
@@ -434,6 +435,75 @@ class Generator:
         )
         self.schema_builder = SchemaBuilder(schema_generator)
         self.responses = {}
+        
+        # Create standard error responses
+        self._create_standard_error_responses()
+
+    def _create_standard_error_responses(self) -> None:
+        """
+        Creates standard error responses that can be reused across operations.
+        These will be added to the components.responses section of the OpenAPI document.
+        """
+        # Get the Error schema
+        error_schema = self.schema_builder.classdef_to_ref(Error)
+        
+        # Create standard error responses
+        self.responses["BadRequest400"] = Response(
+            description="The request was invalid or malformed",
+            content={
+                "application/json": MediaType(
+                    schema=error_schema,
+                    example={
+                        "status": 400,
+                        "title": "Bad Request",
+                        "detail": "The request was invalid or malformed",
+                    }
+                )
+            }
+        )
+        
+        self.responses["TooManyRequests429"] = Response(
+            description="The client has sent too many requests in a given amount of time",
+            content={
+                "application/json": MediaType(
+                    schema=error_schema,
+                    example={
+                        "status": 429,
+                        "title": "Too Many Requests",
+                        "detail": "You have exceeded the rate limit. Please try again later.",
+                    }
+                )
+            }
+        )
+        
+        self.responses["InternalServerError500"] = Response(
+            description="The server encountered an unexpected error",
+            content={
+                "application/json": MediaType(
+                    schema=error_schema,
+                    example={
+                        "status": 500,
+                        "title": "Internal Server Error",
+                        "detail": "An unexpected error occurred. Our team has been notified.",
+                    }
+                )
+            }
+        )
+        
+        # Add a default error response for any unhandled error cases
+        self.responses["DefaultError"] = Response(
+            description="An unexpected error occurred",
+            content={
+                "application/json": MediaType(
+                    schema=error_schema,
+                    example={
+                        "status": 0,
+                        "title": "Error",
+                        "detail": "An unexpected error occurred",
+                    }
+                )
+            }
+        )
 
     def _build_type_tag(self, ref: str, schema: Schema) -> Tag:
         # Don't include schema definition in the tag description because for one,
@@ -649,6 +719,18 @@ class Generator:
             responses.update(response_builder.build_response(response_options))
 
         assert len(responses.keys()) > 0, f"No responses found for {op.name}"
+        
+        # Add standard error response references
+        if self.options.include_standard_error_responses:
+            if "400" not in responses:
+                responses["400"] = ResponseRef("BadRequest400")
+            if "429" not in responses:
+                responses["429"] = ResponseRef("TooManyRequests429")
+            if "500" not in responses:
+                responses["500"] = ResponseRef("InternalServerError500")
+            if "default" not in responses:
+                responses["default"] = ResponseRef("DefaultError")
+        
         if op.event_type is not None:
             builder = ContentBuilder(self.schema_builder)
             callbacks = {

docs/openapi_generator/pyopenapi/options.py

@@ -35,6 +35,7 @@ class Options:
     :param error_wrapper: True if errors are encapsulated in an error object wrapper.
     :param property_description_fun: Custom transformation function to apply to class property documentation strings.
     :param captions: User-defined captions for sections such as "Operations" or "Types", and (if applicable) groups of extra types.
+    :param include_standard_error_responses: Whether to include standard error responses (400, 429, 500, 503) in all operations.
     """
 
     server: Server
@@ -52,6 +53,7 @@ class Options:
     error_wrapper: bool = False
     property_description_fun: Optional[Callable[[type, str, str], str]] = None
     captions: Optional[Dict[str, str]] = None
+    include_standard_error_responses: bool = True
 
     default_captions: ClassVar[Dict[str, str]] = {
         "Operations": "Operations",

docs/source/distributions/building_distro.md

@@ -106,7 +106,7 @@ It would be best to start with a template and understand the structure of the co
 llama stack build
 
 > Enter a name for your Llama Stack (e.g. my-local-stack): my-stack
-> Enter the image type you want your Llama Stack to be built as (container or conda): conda
+> Enter the image type you want your Llama Stack to be built as (container or conda or venv): conda
 
 Llama Stack is composed of several APIs working together. Let's select
 the provider types (implementations) you want to use for these APIs.
@@ -187,7 +187,7 @@ usage: llama stack run [-h] [--port PORT] [--image-name IMAGE_NAME] [--disable-i
                        [--tls-certfile TLS_CERTFILE] [--image-type {conda,container,venv}]
                        config
 
-start the server for a Llama Stack Distribution. You should have already built (or downloaded) and configured the distribution.
+Start the server for a Llama Stack Distribution. You should have already built (or downloaded) and configured the distribution.
 
 positional arguments:
   config                Path to config file to use for the run

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

@@ -41,12 +41,31 @@ The following environment variables can be configured:
 
 ## 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/references/llama_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.
+Please use `llama model list --downloaded` to check that you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/references/llama_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
+$ llama model list --downloaded
-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
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
 ```
 
 ## Running the Distribution

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

@@ -41,12 +41,31 @@ The following environment variables can be configured:
 
 ## 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/references/llama_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.
+Please use `llama model list --downloaded` to check that you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/references/llama_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
+$ llama model list --downloaded
-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
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
 ```
 
 ## Running the Distribution

docs/source/references/llama_cli_reference/download_models.md

@@ -129,3 +129,35 @@ llama download --source huggingface --model-id Prompt-Guard-86M --ignore-pattern
 **Important:** Set your environment variable `HF_TOKEN` or pass in `--hf-token` to the command to validate your access. You can find your token at [https://huggingface.co/settings/tokens](https://huggingface.co/settings/tokens).
 
 > **Tip:** Default for `llama download` is to run with `--ignore-patterns *.safetensors` since we use the `.pth` files in the `original` folder. For Llama Guard and Prompt Guard, however, we need safetensors. Hence, please run with `--ignore-patterns original` so that safetensors are downloaded and `.pth` files are ignored.
+
+## List the downloaded models
+
+To list the downloaded models with the following command:
+```
+llama model list --downloaded
+```
+
+You should see a table like this:
+```
+┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━┓
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
+```

docs/source/references/llama_cli_reference/index.md

@@ -154,6 +154,38 @@ llama download --source huggingface --model-id Prompt-Guard-86M --ignore-pattern
 
 > **Tip:** Default for `llama download` is to run with `--ignore-patterns *.safetensors` since we use the `.pth` files in the `original` folder. For Llama Guard and Prompt Guard, however, we need safetensors. Hence, please run with `--ignore-patterns original` so that safetensors are downloaded and `.pth` files are ignored.
 
+## List the downloaded models
+
+To list the downloaded models with the following command:
+```
+llama model list --downloaded
+```
+
+You should see a table like this:
+```
+┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━┓
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
+```
+
 
 ## Understand the models
 The `llama model` command helps you explore the model’s interface.

docs/zero_to_hero_guide/README.md

@@ -73,7 +73,7 @@ If you're looking for more specific topics, we have a [Zero to Hero Guide](#next
    Open a new terminal and install `llama-stack`:
    ```bash
    conda activate ollama
-   pip install llama-stack==0.1.0
+   pip install -U llama-stack
    ```
 
 ---

llama_stack/apis/datatypes.py

@@ -5,6 +5,9 @@
 # the root directory of this source tree.
 
 from enum import Enum
+from typing import Optional
+
+from pydantic import BaseModel
 
 from llama_stack.schema_utils import json_schema_type
 
@@ -33,3 +36,20 @@ class Api(Enum):
 
     # built-in API
     inspect = "inspect"
+
+
+@json_schema_type
+class Error(BaseModel):
+    """
+    Error response from the API. Roughly follows RFC 7807.
+
+    :param status: HTTP status code
+    :param title: Error title, a short summary of the error which is invariant for an error type
+    :param detail: Error detail, a longer human-readable description of the error
+    :param instance: (Optional) A URL which can be used to retrieve more information about the specific occurrence of the error
+    """
+
+    status: int
+    title: str
+    detail: str
+    instance: Optional[str] = None

llama_stack/distribution/build.py

@@ -15,7 +15,6 @@ from termcolor import cprint
 
 from llama_stack.distribution.datatypes import BuildConfig, Provider
 from llama_stack.distribution.distribution import get_provider_registry
-from llama_stack.distribution.utils.config_dirs import BUILDS_BASE_DIR
 from llama_stack.distribution.utils.exec import run_command, run_with_pty
 from llama_stack.distribution.utils.image_types import ImageType
 from llama_stack.providers.datatypes import Api
@@ -103,8 +102,6 @@ def build_image(
             template_or_config,
             image_name,
             container_base,
-            str(build_file_path),
-            str(BUILDS_BASE_DIR / ImageType.container.value),
             " ".join(normal_deps),
         ]
     elif build_config.image_type == ImageType.conda.value:

llama_stack/distribution/build_conda_env.sh

@@ -52,7 +52,7 @@ ensure_conda_env_python310() {
   local python_version="3.10"
 
   # Check if conda command is available
-  if ! command -v conda &>/dev/null; then
+  if ! is_command_available conda; then
     printf "${RED}Error: conda command not found. Is Conda installed and in your PATH?${NC}" >&2
     exit 1
   fi

llama_stack/distribution/build_container.sh

@@ -1,4 +1,4 @@
-#!/bin/bash
+#!/usr/bin/env bash
 
 # Copyright (c) Meta Platforms, Inc. and affiliates.
 # All rights reserved.
@@ -20,26 +20,27 @@ UV_HTTP_TIMEOUT=${UV_HTTP_TIMEOUT:-500}
 # mounting is not supported by docker buildx, so we use COPY instead
 USE_COPY_NOT_MOUNT=${USE_COPY_NOT_MOUNT:-}
 
-if [ "$#" -lt 6 ]; then
+if [ "$#" -lt 4 ]; then
   # This only works for templates
-  echo "Usage: $0 <template_or_config> <image_name> <container_base> <build_file_path> <host_build_dir> <pip_dependencies> [<special_pip_deps>]" >&2
+  echo "Usage: $0 <template_or_config> <image_name> <container_base> <pip_dependencies> [<special_pip_deps>]" >&2
   exit 1
 fi
 
 set -euo pipefail
 
 template_or_config="$1"
-image_name="$2"
+shift
-container_base="$3"
+image_name="$1"
-build_file_path="$4"
+shift
-host_build_dir="$5"
+container_base="$1"
-pip_dependencies="$6"
+shift
-special_pip_deps="${7:-}"
+pip_dependencies="$1"
+shift
+special_pip_deps="${1:-}"
 
 
 # Define color codes
 RED='\033[0;31m'
-GREEN='\033[0;32m'
 NC='\033[0m' # No Color
 
 CONTAINER_BINARY=${CONTAINER_BINARY:-docker}
@@ -47,8 +48,10 @@ CONTAINER_OPTS=${CONTAINER_OPTS:-}
 
 TEMP_DIR=$(mktemp -d)
 
+SCRIPT_DIR=$(dirname "$(readlink -f "$0")")
+source "$SCRIPT_DIR/common.sh"
+
 add_to_container() {
-  local input
   output_file="$TEMP_DIR/Containerfile"
   if [ -t 0 ]; then
     printf '%s\n' "$1" >>"$output_file"
@@ -58,15 +61,21 @@ add_to_container() {
   fi
 }
 
+# Check if container command is available
+if ! is_command_available $CONTAINER_BINARY; then
+  printf "${RED}Error: ${CONTAINER_BINARY} command not found. Is ${CONTAINER_BINARY} installed and in your PATH?${NC}" >&2
+  exit 1
+fi
+
 # Update and install UBI9 components if UBI9 base image is used
 if [[ $container_base == *"registry.access.redhat.com/ubi9"* ]]; then
   add_to_container << EOF
 FROM $container_base
 WORKDIR /app
 
-RUN microdnf -y update && microdnf install -y iputils net-tools wget \
+RUN dnf -y update && dnf install -y iputils net-tools wget \
     vim-minimal python3.11 python3.11-pip python3.11-wheel \
-    python3.11-setuptools && ln -s /bin/pip3.11 /bin/pip && ln -s /bin/python3.11 /bin/python && microdnf clean all
+    python3.11-setuptools && ln -s /bin/pip3.11 /bin/pip && ln -s /bin/python3.11 /bin/python && dnf clean all
 
 ENV UV_SYSTEM_PYTHON=1
 RUN pip install uv
@@ -150,12 +159,12 @@ EOF
     add_to_container << EOF
 RUN uv pip install --no-cache --extra-index-url https://test.pypi.org/simple/ \
   --index-strategy unsafe-best-match \
-  llama-models==$TEST_PYPI_VERSION llama-stack-client==$TEST_PYPI_VERSION llama-stack==$TEST_PYPI_VERSION
+  llama-stack==$TEST_PYPI_VERSION
 
 EOF
   else
     if [ -n "$PYPI_VERSION" ]; then
-      SPEC_VERSION="llama-stack==${PYPI_VERSION} llama-models==${PYPI_VERSION} llama-stack-client==${PYPI_VERSION}"
+      SPEC_VERSION="llama-stack==${PYPI_VERSION}"
     else
       SPEC_VERSION="llama-stack"
     fi
@@ -165,6 +174,11 @@ EOF
   fi
 fi
 
+# remove uv after installation
+  add_to_container << EOF
+RUN pip uninstall -y uv
+EOF
+
 # if template_or_config ends with .yaml, it is not a template and we should not use the --template flag
 if [[ "$template_or_config" != *.yaml ]]; then
   add_to_container << EOF
@@ -185,26 +199,31 @@ RUN mkdir -p /.llama /.cache
 RUN chmod -R g+rw /app /.llama /.cache
 EOF
 
-printf "Containerfile created successfully in $TEMP_DIR/Containerfile\n\n"
+printf "Containerfile created successfully in %s/Containerfile\n\n" "$TEMP_DIR"
-cat $TEMP_DIR/Containerfile
+cat "$TEMP_DIR"/Containerfile
 printf "\n"
 
-mounts=""
+# Start building the CLI arguments
+CLI_ARGS=()
+
+# Read CONTAINER_OPTS and put it in an array
+read -ra CLI_ARGS <<< "$CONTAINER_OPTS"
+
 if [ "$USE_COPY_NOT_MOUNT" != "true" ]; then
   if [ -n "$LLAMA_STACK_DIR" ]; then
-    mounts="$mounts -v $(readlink -f $LLAMA_STACK_DIR):$stack_mount"
+    CLI_ARGS+=("-v" "$(readlink -f "$LLAMA_STACK_DIR"):$stack_mount")
   fi
   if [ -n "$LLAMA_MODELS_DIR" ]; then
-    mounts="$mounts -v $(readlink -f $LLAMA_MODELS_DIR):$models_mount"
+    CLI_ARGS+=("-v" "$(readlink -f "$LLAMA_MODELS_DIR"):$models_mount")
   fi
   if [ -n "$LLAMA_STACK_CLIENT_DIR" ]; then
-    mounts="$mounts -v $(readlink -f $LLAMA_STACK_CLIENT_DIR):$client_mount"
+    CLI_ARGS+=("-v" "$(readlink -f "$LLAMA_STACK_CLIENT_DIR"):$client_mount")
   fi
 fi
 
-if command -v selinuxenabled &>/dev/null && selinuxenabled; then
+if is_command_available selinuxenabled && selinuxenabled; then
   # Disable SELinux labels -- we don't want to relabel the llama-stack source dir
-  CONTAINER_OPTS="$CONTAINER_OPTS --security-opt label=disable"
+  CLI_ARGS+=("--security-opt" "label=disable")
 fi
 
 # Set version tag based on PyPI version
@@ -225,11 +244,11 @@ image_tag="$image_name:$version_tag"
 # Detect platform architecture
 ARCH=$(uname -m)
 if [ -n "$BUILD_PLATFORM" ]; then
-  PLATFORM="--platform $BUILD_PLATFORM"
+  CLI_ARGS+=("--platform $BUILD_PLATFORM")
 elif [ "$ARCH" = "arm64" ] || [ "$ARCH" = "aarch64" ]; then
-  PLATFORM="--platform linux/arm64"
+  CLI_ARGS+=("--platform" "linux/arm64")
 elif [ "$ARCH" = "x86_64" ]; then
-  PLATFORM="--platform linux/amd64"
+  CLI_ARGS+=("--platform" "linux/amd64")
 else
   echo "Unsupported architecture: $ARCH"
   exit 1
@@ -238,8 +257,13 @@ fi
 echo "PWD: $(pwd)"
 echo "Containerfile: $TEMP_DIR/Containerfile"
 set -x
-$CONTAINER_BINARY build $CONTAINER_OPTS $PLATFORM -t $image_tag \
+
-  -f "$TEMP_DIR/Containerfile" "." $mounts --progress=plain
+$CONTAINER_BINARY build \
+  "${CLI_ARGS[@]}" \
+  -t "$image_tag" \
+  -f "$TEMP_DIR/Containerfile" \
+  "." \
+  --progress=plain
 
 # clean up tmp/configs
 set +x

llama_stack/providers/inline/eval/meta_reference/eval.py

@@ -3,6 +3,7 @@
 #
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
+import json
 from typing import Any, Dict, List, Optional
 
 from tqdm import tqdm
@@ -116,7 +117,7 @@ class MetaReferenceEvalImpl(
         generations = []
         for i, x in tqdm(enumerate(input_rows)):
             assert ColumnName.chat_completion_input.value in x, "Invalid input row"
-            input_messages = eval(str(x[ColumnName.chat_completion_input.value]))
+            input_messages = json.loads(x[ColumnName.chat_completion_input.value])
             input_messages = [UserMessage(**x) for x in input_messages]
 
             # NOTE: only single-turn agent generation is supported. Create a new session for each input row
@@ -158,7 +159,7 @@ class MetaReferenceEvalImpl(
         generations = []
         for x in tqdm(input_rows):
             if ColumnName.completion_input.value in x:
-                input_content = eval(str(x[ColumnName.completion_input.value]))
+                input_content = json.loads(x[ColumnName.completion_input.value])
                 response = await self.inference_api.completion(
                     model=candidate.model,
                     content=input_content,
@@ -166,9 +167,8 @@ class MetaReferenceEvalImpl(
                 )
                 generations.append({ColumnName.generated_answer.value: response.completion_message.content})
             elif ColumnName.chat_completion_input.value in x:
-                chat_completion_input_str = str(x[ColumnName.chat_completion_input.value])
+                chat_completion_input_json = json.loads(x[ColumnName.chat_completion_input.value])
-                input_messages = eval(chat_completion_input_str)
+                input_messages = [UserMessage(**x) for x in chat_completion_input_json]
-                input_messages = [UserMessage(**x) for x in input_messages]
                 messages = []
                 if candidate.system_message:
                     messages.append(candidate.system_message)

llama_stack/providers/inline/post_training/torchtune/datasets/format_adapter.py

@@ -10,16 +10,19 @@
 # This source code is licensed under the BSD-style license found in the
 # LICENSE file in the root directory of this source tree.
 
+import json
 from typing import Any, Mapping
 
 from llama_stack.providers.utils.common.data_schema_validator import ColumnName
 
 
-def llama_stack_instruct_to_torchtune_instruct(sample: Mapping[str, Any]) -> Mapping[str, Any]:
+def llama_stack_instruct_to_torchtune_instruct(
+    sample: Mapping[str, Any],
+) -> Mapping[str, Any]:
     assert ColumnName.chat_completion_input.value in sample and ColumnName.expected_answer.value in sample, (
         "Invalid input row"
     )
-    input_messages = eval(str(sample[ColumnName.chat_completion_input.value]))
+    input_messages = json.loads(sample[ColumnName.chat_completion_input.value])
 
     assert len(input_messages) == 1, "llama stack intruct dataset format only supports 1 user message"
     input_message = input_messages[0]
@@ -37,7 +40,7 @@ def llama_stack_instruct_to_torchtune_instruct(sample: Mapping[str, Any]) -> Map
 def llama_stack_chat_to_torchtune_chat(sample: Mapping[str, Any]) -> Mapping[str, Any]:
     assert ColumnName.dialog.value in sample, "Invalid input row"
     role_map = {"user": "human", "assistant": "gpt"}
-    dialog = eval(str(sample[ColumnName.dialog.value]))
+    dialog = json.loads(sample[ColumnName.dialog.value])
 
     assert len(dialog) > 1, "dialog must have at least 2 messagse"
     roles = []

llama_stack/providers/remote/inference/vllm/vllm.py

@@ -270,6 +270,12 @@ class VLLMInferenceAdapter(Inference, ModelsProtocolPrivate):
         tool_config: Optional[ToolConfig] = None,
     ) -> AsyncGenerator:
         model = await self.model_store.get_model(model_id)
+        # This is to be consistent with OpenAI API and support vLLM <= v0.6.3
+        # References:
+        #   * https://platform.openai.com/docs/api-reference/chat/create#chat-create-tool_choice
+        #   * https://github.com/vllm-project/vllm/pull/10000
+        if not tools and tool_config is not None:
+            tool_config.tool_choice = ToolChoice.none
         request = ChatCompletionRequest(
             model=model.provider_resource_id,
             messages=messages,

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

@@ -29,12 +29,31 @@ The following environment variables can be configured:
 
 ## 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/references/llama_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.
+Please use `llama model list --downloaded` to check that you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/references/llama_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
+$ llama model list --downloaded
-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
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
 ```
 
 ## Running the Distribution

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

@@ -31,12 +31,31 @@ The following environment variables can be configured:
 
 ## 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/references/llama_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.
+Please use `llama model list --downloaded` to check that you have llama model checkpoints downloaded in `~/.llama` before proceeding. See [installation guide](https://llama-stack.readthedocs.io/en/latest/references/llama_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
+$ llama model list --downloaded
-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
+┃ Model                                   ┃ Size     ┃ Modified Time       ┃
+┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━┩
+│ Llama3.2-1B-Instruct:int4-qlora-eo8     │ 1.53 GB  │ 2025-02-26 11:22:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B                             │ 2.31 GB  │ 2025-02-18 21:48:52 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Prompt-Guard-86M                        │ 0.02 GB  │ 2025-02-26 11:29:28 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B-Instruct:int4-spinquant-eo8 │ 3.69 GB  │ 2025-02-26 11:37:41 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-3B                             │ 5.99 GB  │ 2025-02-18 21:51:26 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.1-8B                             │ 14.97 GB │ 2025-02-16 10:36:37 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama3.2-1B-Instruct:int4-spinquant-eo8 │ 1.51 GB  │ 2025-02-26 11:35:02 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B                        │ 2.80 GB  │ 2025-02-26 11:20:46 │
+├─────────────────────────────────────────┼──────────┼─────────────────────┤
+│ Llama-Guard-3-1B:int4                   │ 0.43 GB  │ 2025-02-26 11:33:33 │
+└─────────────────────────────────────────┴──────────┴─────────────────────┘
 ```
 
 ## Running the Distribution

tests/client-sdk/agents/test_agents.py

@@ -4,20 +4,15 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
-import json
-from typing import Dict, List
 from uuid import uuid4
 
 import pytest
 from llama_stack_client.lib.agents.agent import Agent
-from llama_stack_client.lib.agents.client_tool import ClientTool
+from llama_stack_client.lib.agents.client_tool import client_tool
 from llama_stack_client.lib.agents.event_logger import EventLogger
-from llama_stack_client.types import ToolResponseMessage
 from llama_stack_client.types.agents.turn_create_params import Document as AgentDocument
 from llama_stack_client.types.memory_insert_params import Document
-from llama_stack_client.types.shared.completion_message import CompletionMessage
 from llama_stack_client.types.shared_params.agent_config import AgentConfig, ToolConfig
-from llama_stack_client.types.tool_def_param import Parameter
 
 from llama_stack.apis.agents.agents import (
     AgentConfig as Server__AgentConfig,
@@ -27,63 +22,22 @@ from llama_stack.apis.agents.agents import (
 )
 
 
-class TestClientTool(ClientTool):
+@client_tool
-    """Tool to give boiling point of a liquid
+def get_boiling_point(liquid_name: str, celcius: bool = True) -> int:
-    Returns the correct value for polyjuice in Celcius and Fahrenheit
-    and returns -1 for other liquids
     """
+    Returns the boiling point of a liquid in Celcius or Fahrenheit
 
-    def run(self, messages: List[CompletionMessage]) -> List[ToolResponseMessage]:
+    :param liquid_name: The name of the liquid
-        assert len(messages) == 1, "Expected single message"
+    :param celcius: Whether to return the boiling point in Celcius
-
+    :return: The boiling point of the liquid in Celcius or Fahrenheit
-        message = messages[0]
+    """
-
+    if liquid_name.lower() == "polyjuice":
-        tool_call = message.tool_calls[0]
+        if celcius:
-
+            return -100
-        try:
-            response = self.run_impl(**tool_call.arguments)
-            response_str = json.dumps(response, ensure_ascii=False)
-        except Exception as e:
-            response_str = f"Error when running tool: {e}"
-
-        message = ToolResponseMessage(
-            role="tool",
-            call_id=tool_call.call_id,
-            tool_name=tool_call.tool_name,
-            content=response_str,
-        )
-        return message
-
-    def get_name(self) -> str:
-        return "get_boiling_point"
-
-    def get_description(self) -> str:
-        return "Get the boiling point of imaginary liquids (eg. polyjuice)"
-
-    def get_params_definition(self) -> Dict[str, Parameter]:
-        return {
-            "liquid_name": Parameter(
-                name="liquid_name",
-                parameter_type="string",
-                description="The name of the liquid",
-                required=True,
-            ),
-            "celcius": Parameter(
-                name="celcius",
-                parameter_type="boolean",
-                description="Whether to return the boiling point in Celcius",
-                required=False,
-            ),
-        }
-
-    def run_impl(self, liquid_name: str, celcius: bool = True) -> int:
-        if liquid_name.lower() == "polyjuice":
-            if celcius:
-                return -100
-            else:
-                return -212
         else:
-            return -1
+            return -212
+    else:
+        return -1
 
 
 @pytest.fixture(scope="session")
@@ -298,7 +252,7 @@ def test_code_interpreter_for_attachments(llama_stack_client, agent_config):
 
 
 def test_custom_tool(llama_stack_client, agent_config):
-    client_tool = TestClientTool()
+    client_tool = get_boiling_point
     agent_config = {
         **agent_config,
         "toolgroups": ["builtin::websearch"],
@@ -326,7 +280,7 @@ def test_custom_tool(llama_stack_client, agent_config):
 
 def test_tool_choice(llama_stack_client, agent_config):
     def run_agent(tool_choice):
-        client_tool = TestClientTool()
+        client_tool = get_boiling_point
 
         test_agent_config = {
             **agent_config,
@@ -505,7 +459,7 @@ def test_rag_and_code_agent(llama_stack_client, agent_config):
 
 
 def test_create_turn_response(llama_stack_client, agent_config):
-    client_tool = TestClientTool()
+    client_tool = get_boiling_point
     agent_config = {
         **agent_config,
         "input_shields": [],

tests/client-sdk/inference/test_embedding.py

@@ -76,6 +76,25 @@ DUMMY_IMAGE_URL = ImageContentItem(
 )
 DUMMY_IMAGE_BASE64 = ImageContentItem(image=ImageContentItemImage(data="base64string"), type="image")
 SUPPORTED_PROVIDERS = {"remote::nvidia"}
+MODELS_SUPPORTING_MEDIA = {}
+MODELS_SUPPORTING_OUTPUT_DIMENSION = {"nvidia/llama-3.2-nv-embedqa-1b-v2"}
+MODELS_REQUIRING_TASK_TYPE = {
+    "nvidia/llama-3.2-nv-embedqa-1b-v2",
+    "nvidia/nv-embedqa-e5-v5",
+    "nvidia/nv-embedqa-mistral-7b-v2",
+    "snowflake/arctic-embed-l",
+}
+MODELS_SUPPORTING_TASK_TYPE = MODELS_REQUIRING_TASK_TYPE
+
+
+def default_task_type(model_id):
+    """
+    Some models require a task type parameter. This provides a default value for
+    testing those models.
+    """
+    if model_id in MODELS_REQUIRING_TASK_TYPE:
+        return {"task_type": "query"}
+    return {}
 
 
 @pytest.mark.parametrize(
@@ -92,7 +111,9 @@ SUPPORTED_PROVIDERS = {"remote::nvidia"}
 def test_embedding_text(llama_stack_client, embedding_model_id, contents, inference_provider_type):
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
-    response = llama_stack_client.inference.embeddings(model_id=embedding_model_id, contents=contents)
+    response = llama_stack_client.inference.embeddings(
+        model_id=embedding_model_id, contents=contents, **default_task_type(embedding_model_id)
+    )
     assert isinstance(response, EmbeddingsResponse)
     assert len(response.embeddings) == sum(len(content) if isinstance(content, list) else 1 for content in contents)
     assert isinstance(response.embeddings[0], list)
@@ -110,11 +131,14 @@ def test_embedding_text(llama_stack_client, embedding_model_id, contents, infere
         "list[url,string,base64,text]",
     ],
 )
-@pytest.mark.xfail(reason="Media is not supported")
 def test_embedding_image(llama_stack_client, embedding_model_id, contents, inference_provider_type):
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
-    response = llama_stack_client.inference.embeddings(model_id=embedding_model_id, contents=contents)
+    if embedding_model_id not in MODELS_SUPPORTING_MEDIA:
+        pytest.xfail(f"{embedding_model_id} doesn't support media")
+    response = llama_stack_client.inference.embeddings(
+        model_id=embedding_model_id, contents=contents, **default_task_type(embedding_model_id)
+    )
     assert isinstance(response, EmbeddingsResponse)
     assert len(response.embeddings) == sum(len(content) if isinstance(content, list) else 1 for content in contents)
     assert isinstance(response.embeddings[0], list)
@@ -145,7 +169,10 @@ def test_embedding_truncation(
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
     response = llama_stack_client.inference.embeddings(
-        model_id=embedding_model_id, contents=contents, text_truncation=text_truncation
+        model_id=embedding_model_id,
+        contents=contents,
+        text_truncation=text_truncation,
+        **default_task_type(embedding_model_id),
     )
     assert isinstance(response, EmbeddingsResponse)
     assert len(response.embeddings) == 1
@@ -178,26 +205,36 @@ def test_embedding_truncation_error(
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
     with pytest.raises(BadRequestError):
         llama_stack_client.inference.embeddings(
-            model_id=embedding_model_id, contents=[DUMMY_LONG_TEXT], text_truncation=text_truncation
+            model_id=embedding_model_id,
+            contents=[DUMMY_LONG_TEXT],
+            text_truncation=text_truncation,
+            **default_task_type(embedding_model_id),
         )
 
 
-@pytest.mark.xfail(reason="Only valid for model supporting dimension reduction")
 def test_embedding_output_dimension(llama_stack_client, embedding_model_id, inference_provider_type):
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
-    base_response = llama_stack_client.inference.embeddings(model_id=embedding_model_id, contents=[DUMMY_STRING])
+    if embedding_model_id not in MODELS_SUPPORTING_OUTPUT_DIMENSION:
+        pytest.xfail(f"{embedding_model_id} doesn't support output_dimension")
+    base_response = llama_stack_client.inference.embeddings(
+        model_id=embedding_model_id, contents=[DUMMY_STRING], **default_task_type(embedding_model_id)
+    )
     test_response = llama_stack_client.inference.embeddings(
-        model_id=embedding_model_id, contents=[DUMMY_STRING], output_dimension=32
+        model_id=embedding_model_id,
+        contents=[DUMMY_STRING],
+        **default_task_type(embedding_model_id),
+        output_dimension=32,
     )
     assert len(base_response.embeddings[0]) != len(test_response.embeddings[0])
     assert len(test_response.embeddings[0]) == 32
 
 
-@pytest.mark.xfail(reason="Only valid for model supporting task type")
 def test_embedding_task_type(llama_stack_client, embedding_model_id, inference_provider_type):
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
+    if embedding_model_id not in MODELS_SUPPORTING_TASK_TYPE:
+        pytest.xfail(f"{embedding_model_id} doesn't support task_type")
     query_embedding = llama_stack_client.inference.embeddings(
         model_id=embedding_model_id, contents=[DUMMY_STRING], task_type="query"
     )
@@ -220,7 +257,10 @@ def test_embedding_text_truncation(llama_stack_client, embedding_model_id, text_
     if inference_provider_type not in SUPPORTED_PROVIDERS:
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
     response = llama_stack_client.inference.embeddings(
-        model_id=embedding_model_id, contents=[DUMMY_STRING], text_truncation=text_truncation
+        model_id=embedding_model_id,
+        contents=[DUMMY_STRING],
+        text_truncation=text_truncation,
+        **default_task_type(embedding_model_id),
     )
     assert isinstance(response, EmbeddingsResponse)
     assert len(response.embeddings) == 1
@@ -245,5 +285,8 @@ def test_embedding_text_truncation_error(
         pytest.xfail(f"{inference_provider_type} doesn't support embedding model yet")
     with pytest.raises(BadRequestError):
         llama_stack_client.inference.embeddings(
-            model_id=embedding_model_id, contents=[DUMMY_STRING], text_truncation=text_truncation
+            model_id=embedding_model_id,
+            contents=[DUMMY_STRING],
+            text_truncation=text_truncation,
+            **default_task_type(embedding_model_id),
         )