#!/bin/bash # SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. # SPDX-License-Identifier: Apache-2.0 # # Aggregated embedding model serving. # GPUs: 1 set -e trap 'echo Cleaning up...; kill 0' EXIT SCRIPT_DIR="$(dirname "$(readlink -f "$0")")" source "$SCRIPT_DIR/../../../common/gpu_utils.sh" # build_gpu_mem_args source "$SCRIPT_DIR/../../../common/launch_utils.sh" # print_launch_banner, wait_any_exit # Default values MODEL="Qwen/Qwen3-Embedding-4B" # Parse command line arguments EXTRA_ARGS=() while [[ $# -gt 0 ]]; do case $1 in --model-path) MODEL="$2" shift 2 ;; -h|--help) echo "Usage: $0 [OPTIONS]" echo "Options:" echo " --model-path Specify model (default: $MODEL)" echo " -h, --help Show this help message" echo "" echo "Note: System metrics are enabled by default on port 8081 (worker)" echo "Note: OpenTelemetry tracing is not yet supported for embedding models" exit 0 ;; *) EXTRA_ARGS+=("$1") shift ;; esac done GPU_MEM_ARGS=$(build_gpu_mem_args sglang) HTTP_PORT="${DYN_HTTP_PORT:-8000}" print_launch_banner --no-curl "Launching Embedding Worker" "$MODEL" "$HTTP_PORT" print_curl_footer <