Skip to content

Fix duplicate flag addition and make sleep time configurable #20

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Mar 27, 2025
Merged

Fix duplicate flag addition and make sleep time configurable #20

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
40c47e8
first commit
Bslabe123 Mar 27, 2025
ec19c61
dont remove licence
Bslabe123 Mar 27, 2025
d606e2c
revert todo
Bslabe123 Mar 27, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
64 changes: 34 additions & 30 deletions latency_throughput_curve.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

set -o xtrace

export IP=$IP
Expand All @@ -24,48 +25,51 @@ if [[ "$PROMPT_DATASET" = "sharegpt" ]]; then
fi

PYTHON="python3"
PYTHON_OPTS="benchmark_serving.py "
BASE_PYTHON_OPTS=(
"benchmark_serving.py"
"--save-json-results"
"--host=$IP"
"--port=$PORT"
"--dataset=$PROMPT_DATASET_FILE"
"--tokenizer=$TOKENIZER"
"--backend=$BACKEND"
"--max-input-length=$INPUT_LENGTH"
"--max-output-length=$OUTPUT_LENGTH"
"--file-prefix=$FILE_PREFIX"
"--models=$MODELS"
"--pm-namespace=$PM_NAMESPACE"
"--pm-job=$PM_JOB"
)

[[ "$TRAFFIC_SPLIT" ]] && BASE_PYTHON_OPTS+=("--traffic-split=$TRAFFIC_SPLIT")
[[ "$OUTPUT_BUCKET" ]] && BASE_PYTHON_OPTS+=("--output-bucket=$OUTPUT_BUCKET")
[[ "$SCRAPE_SERVER_METRICS" = "true" ]] && BASE_PYTHON_OPTS+=("--scrape-server-metrics")
[[ "$SAVE_AGGREGATED_RESULT" = "true" ]] && BASE_PYTHON_OPTS+=("--save-aggregated-result")
[[ "$STREAM_REQUEST" = "true" ]] && BASE_PYTHON_OPTS+=("--stream-request")
[[ "$OUTPUT_BUCKET_FILEPATH" ]] && BASE_PYTHON_OPTS+=("--output-bucket-filepath" "$OUTPUT_BUCKET_FILEPATH")

SLEEP_TIME=${SLEEP_TIME:-0}

for request_rate in $(echo $REQUEST_RATES | tr ',' ' '); do
echo "Benchmarking request rate: ${request_rate}"
# TODO: Check if profile already exists, if so then skip
timestamp=$(date +"%Y-%m-%d_%H-%M-%S")
output_file="latency-profile-${timestamp}.txt"
if [ ${request_rate} == 0 ]; then

if [ "$request_rate" == "0" ]; then
request_rate="inf"
num_prompts=$MAX_NUM_PROMPTS
else
num_prompts=$(awk "BEGIN {print int($request_rate * $BENCHMARK_TIME_SECONDS)}")
fi

echo "TOTAL prompts: $num_prompts"

# Build the python command options
PYTHON_OPTS="$PYTHON_OPTS --save-json-results --host=$IP --port=$PORT --dataset=$PROMPT_DATASET_FILE --tokenizer=$TOKENIZER --request-rate=$request_rate --backend=$BACKEND --num-prompts=$num_prompts --max-input-length=$INPUT_LENGTH --max-output-length=$OUTPUT_LENGTH --file-prefix=$FILE_PREFIX --models=$MODELS --pm-namespace=$PM_NAMESPACE --pm-job=$PM_JOB"

if [[ "$TRAFFIC_SPLIT" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --traffic-split=$TRAFFIC_SPLIT"
fi

if [[ "$OUTPUT_BUCKET" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --output-bucket=$OUTPUT_BUCKET"
fi

if [[ "$SCRAPE_SERVER_METRICS" = "true" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --scrape-server-metrics"
fi
if [[ "$SAVE_AGGREGATED_RESULT" = "true" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --save-aggregated-result"
fi
if [[ "$STREAM_REQUEST" = "true" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --stream-request"
fi
if [[ "$OUTPUT_BUCKET_FILEPATH" ]]; then
PYTHON_OPTS="$PYTHON_OPTS --output-bucket-filepath $OUTPUT_BUCKET_FILEPATH"
fi
echo "TOTAL prompts: $num_prompts"
PYTHON_OPTS=("${BASE_PYTHON_OPTS[@]}" "--request-rate=$request_rate" "--num-prompts=$num_prompts")

$PYTHON $PYTHON_OPTS > $output_file
cat $output_file
sleep 30 # wait 30 seconds before next run to ensure metrics isolation
$PYTHON "${PYTHON_OPTS[@]}" > "$output_file"
cat "$output_file"
echo "Sleeping for $SLEEP_TIME seconds..."
sleep $SLEEP_TIME
done

export LPG_FINISHED="true"
Expand Down