fix: 调整 vllm 启动参数，修复 DotsOCR 模型注册逻辑

zhch/vllm_daemon.sh

@@ -8,6 +8,7 @@ LOGFILE="$LOGDIR/vllm.log"
 conda activate dots.ocr
 hf_model_path="/home/ubuntu/zhch/dots.ocr/weights/DotsOCR"  # Path to your downloaded model weights
 export PYTHONPATH=$(dirname "$hf_model_path"):$PYTHONPATH
+
 # launch vllm server
 model_name="DotsOCR"
 
@@ -21,13 +22,13 @@ start() {
     nohup vllm serve $hf_model_path \
 		--host 0.0.0.0 \
 		--port 8101 \
-		--gpu-memory-utilization 0.90 \
+		--gpu-memory-utilization 0.85 \
 		--max-log-len 1000 \
 		--trust-remote-code \
 		--max-model-len 65536 \
 		--max-num-batched-tokens 65536 \
 		--uvicorn-log-level info \
-		--limit-mm-per-prompt '{"image": 5}' \
+		--limit-mm-per-prompt '{"image": 3}' \
 		--chat-template-content-format string \
 		--data-parallel-size 3 \
 		--max-num-seqs 32 \
@@ -60,6 +61,17 @@ status() {
     fi
 }
 
+# 修复后的 launch_model_vllm.sh
+echo "🔧 注册 DotsOCR 模型到 vLLM..."
+vllm_path=$(which vllm)
+if ! grep -q "from DotsOCR import modeling_dots_ocr_vllm" "$vllm_path"; then
+    sed -i '/^from vllm\.entrypoints\.cli\.main import main$/a\
+from DotsOCR import modeling_dots_ocr_vllm' "$vllm_path"
+    echo "✅ DotsOCR 模型已注册到 vLLM"
+else
+    echo "✅ DotsOCR 模型已经注册过了"
+fi
+
 case "$1" in
     start)
         start