Bug fix: determine agbench_base_dir based on method of install

and use ENV $VENV_BASE_DIR for venv setup

src/autogluon/bench/Dockerfile

@@ -3,7 +3,6 @@ FROM $AG_BENCH_BASE_IMAGE
 
 ENV DEBIAN_FRONTEND=noninteractive
 ENV RUNNING_IN_DOCKER=true
-ENV AGBENCH_BASE=src/autogluon/bench/
 
 # Install essential packages and Python 3.9
 RUN apt-get update && \
@@ -25,37 +24,41 @@ RUN apt-get install -y python3-pip unzip curl git pciutils && \
 
 # Application-specific steps
 ARG AG_BENCH_VERSION
+ARG AG_BENCH_BASE_DIR
 ARG CDK_DEPLOY_REGION
 ARG FRAMEWORK_PATH
 ARG GIT_URI
 ARG GIT_BRANCH
 ARG AMLB_FRAMEWORK
 ARG AMLB_USER_DIR
 
+ENV AG_BENCH_BASE_DIR=${AG_BENCH_BASE_DIR}
+ENV VENV_BASE_DIR="."
+
 WORKDIR /app/
 
-# Copying necessary files for autogluon-bench package
-COPY . /app/
-COPY ${AGBENCH_BASE}entrypoint.sh /app/
-COPY ${AGBENCH_BASE}custom_configs /app/custom_configs/
+# Copying files for installing autogluon.bench from source, user needs to at the root of autogluon.bench repo as we need pyproject.toml
+COPY . .
+COPY ${AG_BENCH_BASE_DIR}/entrypoint.sh .
+COPY ${AG_BENCH_BASE_DIR}/custom_configs custom_configs/
 
 # check if autogluon.bench version contains "dev" tag
 RUN if echo "$AG_BENCH_VERSION" | grep -q "dev"; then \ 
         # install from local source
-        pip install /app/; \
+        pip install .; \
     else \
         pip install autogluon.bench==$AG_BENCH_VERSION; \
     fi
 
 RUN chmod +x entrypoint.sh \
     && if echo "$FRAMEWORK_PATH" | grep -q -E "tabular|timeseries"; then \
         if [ -n "$AMLB_USER_DIR" ]; then \
-            bash ${AGBENCH_BASE}${FRAMEWORK_PATH}setup.sh $GIT_URI $GIT_BRANCH "/home" $AMLB_FRAMEWORK $AMLB_USER_DIR; \
+            bash ${AG_BENCH_BASE_DIR}/${FRAMEWORK_PATH}/setup.sh $GIT_URI $GIT_BRANCH $VENV_BASE_DIR $AMLB_FRAMEWORK $AMLB_USER_DIR; \
         else \
-            bash ${AGBENCH_BASE}${FRAMEWORK_PATH}setup.sh $GIT_URI $GIT_BRANCH "/home" $AMLB_FRAMEWORK; \
+            bash ${AG_BENCH_BASE_DIR}/${FRAMEWORK_PATH}/setup.sh $GIT_URI $GIT_BRANCH $VENV_BASE_DIR $AMLB_FRAMEWORK; \
         fi; \
     elif echo "$FRAMEWORK_PATH" | grep -q "multimodal"; then \
-        bash ${AGBENCH_BASE}${FRAMEWORK_PATH}setup.sh $GIT_URI $GIT_BRANCH "/home" $AG_BENCH_VERSION; \
+        bash ${AG_BENCH_BASE_DIR}/${FRAMEWORK_PATH}/setup.sh $GIT_URI $GIT_BRANCH $VENV_BASE_DIR $AG_BENCH_VERSION; \
     fi \
     && echo "CDK_DEPLOY_REGION=$CDK_DEPLOY_REGION" >> /etc/environment
 

src/autogluon/bench/cloud/aws/batch_stack/stack.py

@@ -37,10 +37,15 @@ def find_project_root_or_fallback(start_dir: str, root_identifier: str = "pyproj
     return start_dir
 
 
+# when pip installed, project_root is at $site_package_path/src/autogluon/bench agbench_base_dir should be ./
+# when installed from source, project_root is at ./, agbench_base_dir should be ./src/autogluon/bench
 with importlib.resources.path("autogluon.bench", "Dockerfile") as file_path:
     docker_base_dir = os.path.dirname(file_path)
     project_root = find_project_root_or_fallback(docker_base_dir)
     docker_path = os.path.relpath(file_path, project_root)
+    agbench_base_dir = os.path.dirname(docker_path)
+    if agbench_base_dir == "":
+        agbench_base_dir = "."
 
 with importlib.resources.path("autogluon.bench.cloud.aws.batch_stack.lambdas", "lambda_function.py") as file_path:
     lambda_script_dir = os.path.dirname(file_path)
@@ -174,6 +179,8 @@ def __init__(self, scope: Construct, id: str, static_stack: StaticResourceStack,
         # Currently CDK can only push to the default repo aws-cdk/assets
         # https://github.com/aws/aws-cdk/issues/12597
         # TODO: use https://github.com/cdklabs/cdk-docker-image-deployment
+
+        logger.info(f"Building Dockerfile at {docker_path} with context at {project_root}")
         docker_image_asset = ecr_assets.DockerImageAsset(
             self,
             f"{prefix}-ecr-docker-image-asset",
@@ -182,6 +189,7 @@ def __init__(self, scope: Construct, id: str, static_stack: StaticResourceStack,
             follow_symlinks=core.SymlinkFollowMode.ALWAYS,
             build_args={
                 "AG_BENCH_BASE_IMAGE": os.environ["AG_BENCH_BASE_IMAGE"],
+                "AG_BENCH_BASE_DIR": agbench_base_dir,
                 "AG_BENCH_VERSION": os.getenv("AG_BENCH_VERSION", "latest"),
                 "CDK_DEPLOY_REGION": os.environ["CDK_DEPLOY_REGION"],
                 "FRAMEWORK_PATH": os.environ["FRAMEWORK_PATH"],

src/autogluon/bench/entrypoint.sh

@@ -1,6 +1,5 @@
 #!/bin/bash
 
 echo "Running hardware utilization monitoring in the background..."
-${AGBENCH_BASE}utils/hardware_utilization.sh &
-
+${AG_BENCH_BASE_DIR}/utils/hardware_utilization.sh &
 agbench run $config_file --skip-setup

src/autogluon/bench/frameworks/multimodal/multimodal_benchmark.py

@@ -89,12 +89,13 @@ def run(
             None
         """
         if os.environ.get("RUNNING_IN_DOCKER", False):
-            venv_base_dir = "/home/"
+            venv_base_dir = os.environ["VENV_BASE_DIR"]
         else:
             venv_base_dir = self.benchmark_dir
         PY_EXC_PATH = os.path.join(venv_base_dir, ".venv/bin/python")
 
-        exec_path = os.path.abspath(os.path.dirname(__file__)) + "/exec.py"
+        exec_path = os.path.join(os.path.abspath(os.path.dirname(__file__)), "exec.py")
+        logger.info(f"Executing {exec_path} under {PY_EXC_PATH}")
         command = [
             PY_EXC_PATH,
             exec_path,

src/autogluon/bench/frameworks/tabular/exec.sh

@@ -30,7 +30,6 @@ if [ -n "$fold" ]; then
 fi
 
 if [ -n "$user_dir" ]; then
-    cp -r $user_dir $venv_base_dir
     amlb_args+=" -u $user_dir"
 fi
 

src/autogluon/bench/frameworks/tabular/tabular_benchmark.py

@@ -53,7 +53,7 @@ def run(
         """
 
         if os.environ.get("RUNNING_IN_DOCKER", False):
-            venv_base_dir = "/home/"
+            venv_base_dir = os.environ["VENV_BASE_DIR"]
         else:
             venv_base_dir = self.benchmark_dir
 

src/autogluon/bench/frameworks/timeseries/exec.sh

@@ -30,7 +30,6 @@ if [ -n "$fold" ]; then
 fi
 
 if [ -n "$user_dir" ]; then
-    cp -r $user_dir $venv_base_dir
     amlb_args+=" -u $user_dir"
 fi
 

src/autogluon/bench/frameworks/timeseries/timeseries_benchmark.py

@@ -51,7 +51,7 @@ def run(
             None
         """
         if os.environ.get("RUNNING_IN_DOCKER", False):
-            venv_base_dir = "/home/"
+            venv_base_dir = os.environ["VENV_BASE_DIR"]
         else:
             venv_base_dir = self.benchmark_dir
 

src/autogluon/bench/runbenchmark.py

@@ -398,7 +398,7 @@ def run(
             # setting environment variables for docker build ARG
             os.environ["AG_BENCH_VERSION"] = agbench_version
 
-            os.environ["FRAMEWORK_PATH"] = f"frameworks/{module}/"
+            os.environ["FRAMEWORK_PATH"] = f"frameworks/{module}"
 
             if module in AMLB_DEPENDENT_MODULES:
                 os.environ["AMLB_FRAMEWORK"] = configs["framework"]