test.sh

# Copyright (c) 2022-2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
#
# See LICENSE for license information.

function error_exit() {
    echo "Error: $1"
    exit 1
}

function test_fail() {
    RET=1
    FAILED_CASES="$FAILED_CASES $1"
    echo "Error: sub-test failed: $1"
}

RET=0
FAILED_CASES=""

: ${TE_PATH:=/opt/transformerengine}
: ${NVTE_TEST_NVINSPECT_FEATURE_DIRS:=$TE_PATH/transformer_engine/debug/features}
: ${NVTE_TEST_NVINSPECT_CONFIGS_DIR:=$TE_PATH/tests/pytorch/debug/test_configs/}
: ${XML_LOG_DIR:=/logs}
mkdir -p "$XML_LOG_DIR"

# Config with the dummy feature which prevents nvinspect from being disabled.
# Nvinspect will be disabled if no feature is active.
: ${NVTE_TEST_NVINSPECT_DUMMY_CONFIG_FILE:=$TE_PATH/tests/pytorch/debug/test_configs/dummy_feature.yaml}

# It is not installed as a requirement,
# because it is not available on PyPI.
pip uninstall -y nvdlfw-inspect
pip install git+https://github.com/NVIDIA/nvidia-dlfw-inspect.git

pip install pytest==8.2.1 || error_exit "Failed to install pytest"

pytest -v -s --junitxml=$XML_LOG_DIR/test_sanity.xml $TE_PATH/tests/pytorch/debug/test_sanity.py  --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS || test_fail "test_sanity.py"
pytest -v -s --junitxml=$XML_LOG_DIR/test_config.xml $TE_PATH/tests/pytorch/debug/test_config.py --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS || test_fail "test_config.py"
pytest -v -s --junitxml=$XML_LOG_DIR/test_numerics.xml $TE_PATH/tests/pytorch/debug/test_numerics.py --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS || test_fail "test_numerics.py"
pytest -v -s --junitxml=$XML_LOG_DIR/test_log.xml $TE_PATH/tests/pytorch/debug/test_log.py --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS --configs_dir=$NVTE_TEST_NVINSPECT_CONFIGS_DIR || test_fail "test_log.py"
NVTE_TORCH_COMPILE=0 pytest -v -s --junitxml=$XML_LOG_DIR/test_api_features.xml $TE_PATH/tests/pytorch/debug/test_api_features.py --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS --configs_dir=$NVTE_TEST_NVINSPECT_CONFIGS_DIR || test_fail "test_api_features.py"
pytest -v -s --junitxml=$XML_LOG_DIR/test_perf.xml $TE_PATH/tests/pytorch/debug/test_perf.py --feature_dirs=$NVTE_TEST_NVINSPECT_FEATURE_DIRS --configs_dir=$NVTE_TEST_NVINSPECT_CONFIGS_DIR || test_fail "test_perf.py"

# standard sanity and numerics tests with initialized debug
NVTE_TEST_NVINSPECT_ENABLED=1 NVTE_TEST_NVINSPECT_CONFIG_FILE=$NVTE_TEST_NVINSPECT_DUMMY_CONFIG_FILE NVTE_TEST_NVINSPECT_FEATURE_DIRS=$NVTE_TEST_NVINSPECT_FEATURE_DIRS PYTORCH_JIT=0 NVTE_TORCH_COMPILE=0 NVTE_ALLOW_NONDETERMINISTIC_ALGO=0 pytest -v -s --junitxml=$XML_LOG_DIR/test_sanity_2.xml $TE_PATH/tests/pytorch/test_sanity.py || test_fail "debug test_sanity.py"
NVTE_TEST_NVINSPECT_ENABLED=1 NVTE_TEST_NVINSPECT_CONFIG_FILE=$NVTE_TEST_NVINSPECT_DUMMY_CONFIG_FILE NVTE_TEST_NVINSPECT_FEATURE_DIRS=$NVTE_TEST_NVINSPECT_FEATURE_DIRS PYTORCH_JIT=0 NVTE_TORCH_COMPILE=0 NVTE_ALLOW_NONDETERMINISTIC_ALGO=0 NVTE_FUSED_ATTN=0 pytest -v -s --junitxml=$XML_LOG_DIR/test_numerics_2.xml $TE_PATH/tests/pytorch/test_numerics.py || test_fail "debug test_numerics.py"

if [ "$RET" -ne 0 ]; then
    echo "Error in the following test cases:$FAILED_CASES"
    exit 1
fi
echo "All tests passed"
exit 0