Use uv pip install instead of pip install (#921)

## What does this PR do? 

See issue: #747 -- `uv` is just plain better. This PR does the bare
minimum of replacing `pip install` by `uv pip install` and ensuring `uv`
exists in the environment.

## Test Plan 

First: create new conda, `uv pip install -e .` on `llama-stack` -- all
is good.
Next: run `llama stack build --template together` followed by `llama
stack run together` -- all good
Next: run `llama stack build --template together --image-name yoyo`
followed by `llama stack run together --image-name yoyo` -- all good
Next: fresh conda and `uv pip install -e .` and `llama stack build
--template together --image-type venv` -- all good.

Docker: `llama stack build --template together --image-type container`
works!

llama_stack/distribution/build.py

@@ -95,11 +95,11 @@ def print_pip_install_help(providers: Dict[str, List[Provider]]):
     normal_deps, special_deps = get_provider_dependencies(providers)
 
     cprint(
-        f"Please install needed dependencies using the following commands:\n\npip install {' '.join(normal_deps)}",
+        f"Please install needed dependencies using the following commands:\n\nuv pip install {' '.join(normal_deps)}",
         "yellow",
     )
     for special_dep in special_deps:
-        cprint(f"pip install {special_dep}", "yellow")
+        cprint(f"uv pip install {special_dep}", "yellow")
     print()
 
 

llama_stack/distribution/build_conda_env.sh

@@ -78,10 +78,12 @@ ensure_conda_env_python310() {
   eval "$(conda shell.bash hook)"
   conda deactivate && conda activate "${env_name}"
 
+  $CONDA_PREFIX/bin/pip install uv
+
   if [ -n "$TEST_PYPI_VERSION" ]; then
     # these packages are damaged in test-pypi, so install them first
-    $CONDA_PREFIX/bin/pip install fastapi libcst
-    $CONDA_PREFIX/bin/pip install --extra-index-url https://test.pypi.org/simple/ \
+    uv pip install fastapi libcst
+    uv pip install --extra-index-url https://test.pypi.org/simple/ \
       llama-models==$TEST_PYPI_VERSION \
       llama-stack-client==$TEST_PYPI_VERSION \
       llama-stack==$TEST_PYPI_VERSION \
@@ -90,7 +92,7 @@ ensure_conda_env_python310() {
       IFS='#' read -ra parts <<<"$special_pip_deps"
       for part in "${parts[@]}"; do
         echo "$part"
-        $CONDA_PREFIX/bin/pip install $part
+        uv pip install $part
       done
     fi
   else
@@ -102,7 +104,7 @@ ensure_conda_env_python310() {
       fi
 
       printf "Installing from LLAMA_STACK_DIR: $LLAMA_STACK_DIR\n"
-      $CONDA_PREFIX/bin/pip install --no-cache-dir -e "$LLAMA_STACK_DIR"
+      uv pip install --no-cache-dir -e "$LLAMA_STACK_DIR"
     else
       PYPI_VERSION="${PYPI_VERSION:-}"
       if [ -n "$PYPI_VERSION" ]; then
@@ -110,7 +112,7 @@ ensure_conda_env_python310() {
       else
         SPEC_VERSION="llama-stack"
       fi
-      $CONDA_PREFIX/bin/pip install --no-cache-dir $SPEC_VERSION
+      uv pip install --no-cache-dir $SPEC_VERSION
     fi
 
     if [ -n "$LLAMA_MODELS_DIR" ]; then
@@ -120,18 +122,18 @@ ensure_conda_env_python310() {
       fi
 
       printf "Installing from LLAMA_MODELS_DIR: $LLAMA_MODELS_DIR\n"
-      $CONDA_PREFIX/bin/pip uninstall -y llama-models
-      $CONDA_PREFIX/bin/pip install --no-cache-dir -e "$LLAMA_MODELS_DIR"
+      uv pip uninstall -y llama-models
+      uv pip install --no-cache-dir -e "$LLAMA_MODELS_DIR"
     fi
 
     # Install pip dependencies
     printf "Installing pip dependencies\n"
-    $CONDA_PREFIX/bin/pip install $pip_dependencies
+    uv pip install $pip_dependencies
     if [ -n "$special_pip_deps" ]; then
       IFS='#' read -ra parts <<<"$special_pip_deps"
       for part in "${parts[@]}"; do
         echo "$part"
-        $CONDA_PREFIX/bin/pip install $part
+        uv pip install $part
       done
     fi
   fi

llama_stack/distribution/build_container.sh

@@ -76,6 +76,8 @@ RUN apt-get update && apt-get install -y \
        bubblewrap \
        && rm -rf /var/lib/apt/lists/*
 
+ENV UV_SYSTEM_PYTHON=1
+RUN pip install uv
 EOF
 fi
 
@@ -83,7 +85,7 @@ fi
 # so we can reuse layers.
 if [ -n "$pip_dependencies" ]; then
   add_to_container << EOF
-RUN pip install --no-cache $pip_dependencies
+RUN uv pip install --no-cache $pip_dependencies
 EOF
 fi
 
@@ -91,7 +93,7 @@ if [ -n "$special_pip_deps" ]; then
   IFS='#' read -ra parts <<<"$special_pip_deps"
   for part in "${parts[@]}"; do
     add_to_container <<EOF
-RUN pip install --no-cache $part
+RUN uv pip install --no-cache $part
 EOF
   done
 fi
@@ -109,16 +111,16 @@ if [ -n "$LLAMA_STACK_DIR" ]; then
   # so that changes will be reflected in the container without having to do a
   # rebuild. This is just for development convenience.
   add_to_container << EOF
-RUN pip install --no-cache -e $stack_mount
+RUN uv pip install --no-cache -e $stack_mount
 EOF
 else
   if [ -n "$TEST_PYPI_VERSION" ]; then
     # these packages are damaged in test-pypi, so install them first
     add_to_container << EOF
-RUN pip install fastapi libcst
+RUN uv pip install fastapi libcst
 EOF
     add_to_container << EOF
-RUN pip install --no-cache --extra-index-url https://test.pypi.org/simple/ \
+RUN uv pip install --no-cache --extra-index-url https://test.pypi.org/simple/ \
   llama-models==$TEST_PYPI_VERSION llama-stack-client==$TEST_PYPI_VERSION llama-stack==$TEST_PYPI_VERSION
 
 EOF
@@ -129,7 +131,7 @@ EOF
       SPEC_VERSION="llama-stack"
     fi
     add_to_container << EOF
-RUN pip install --no-cache $SPEC_VERSION
+RUN uv pip install --no-cache $SPEC_VERSION
 EOF
   fi
 fi
@@ -141,8 +143,8 @@ if [ -n "$LLAMA_MODELS_DIR" ]; then
   fi
 
   add_to_container << EOF
-RUN pip uninstall -y llama-models
-RUN pip install --no-cache $models_mount
+RUN uv pip uninstall -y llama-models
+RUN uv pip install --no-cache $models_mount
 
 EOF
 fi

llama_stack/distribution/build_venv.sh

@@ -51,17 +51,18 @@ run() {
   local pip_dependencies="$2"
   local special_pip_deps="$3"
 
+  pip install uv
   if [ -n "$TEST_PYPI_VERSION" ]; then
     # these packages are damaged in test-pypi, so install them first
-    pip install fastapi libcst
-    pip install --extra-index-url https://test.pypi.org/simple/ \
+    uv pip install fastapi libcst
+    uv pip install --extra-index-url https://test.pypi.org/simple/ \
       llama-models==$TEST_PYPI_VERSION llama-stack==$TEST_PYPI_VERSION \
       $pip_dependencies
     if [ -n "$special_pip_deps" ]; then
       IFS='#' read -ra parts <<<"$special_pip_deps"
       for part in "${parts[@]}"; do
         echo "$part"
-        pip install $part
+        uv pip install $part
       done
     fi
   else
@@ -73,9 +74,9 @@ run() {
       fi
 
       printf "Installing from LLAMA_STACK_DIR: $LLAMA_STACK_DIR\n"
-      pip install --no-cache-dir -e "$LLAMA_STACK_DIR"
+      uv pip install --no-cache-dir -e "$LLAMA_STACK_DIR"
     else
-      pip install --no-cache-dir llama-stack
+      uv pip install --no-cache-dir llama-stack
     fi
 
     if [ -n "$LLAMA_MODELS_DIR" ]; then
@@ -85,18 +86,18 @@ run() {
       fi
 
       printf "Installing from LLAMA_MODELS_DIR: $LLAMA_MODELS_DIR\n"
-      pip uninstall -y llama-models
-      pip install --no-cache-dir -e "$LLAMA_MODELS_DIR"
+      uv pip uninstall -y llama-models
+      uv pip install --no-cache-dir -e "$LLAMA_MODELS_DIR"
     fi
 
     # Install pip dependencies
     printf "Installing pip dependencies\n"
-    pip install $pip_dependencies
+    uv pip install $pip_dependencies
     if [ -n "$special_pip_deps" ]; then
       IFS='#' read -ra parts <<<"$special_pip_deps"
       for part in "${parts[@]}"; do
         echo "$part"
-        pip install $part
+        uv pip install $part
       done
     fi
   fi

llama_stack/distribution/client.py

@@ -68,9 +68,9 @@ def create_api_client_class(protocol) -> Type:
                 return_type = None
             else:
                 return_type = extract_non_async_iterator_type(sig.return_annotation)
-                assert (
-                    return_type
-                ), f"Could not extract return type for {sig.return_annotation}"
+                assert return_type, (
+                    f"Could not extract return type for {sig.return_annotation}"
+                )
 
             async with httpx.AsyncClient() as client:
                 params = self.httpx_request_params(method_name, *args, **kwargs)
@@ -87,9 +87,9 @@ def create_api_client_class(protocol) -> Type:
             webmethod, sig = self.routes[method_name]
 
             return_type = extract_async_iterator_type(sig.return_annotation)
-            assert (
-                return_type
-            ), f"Could not extract return type for {sig.return_annotation}"
+            assert return_type, (
+                f"Could not extract return type for {sig.return_annotation}"
+            )
 
             async with httpx.AsyncClient() as client:
                 params = self.httpx_request_params(method_name, *args, **kwargs)