fix bugs

Author: 11zhouxuan

src/emd/models/engines.py

@@ -75,7 +75,7 @@ class ComfyuiEngine(Engine):
 vllm_deepseek_r1_distill_qwen_engine071 = VllmEngine(**{
             **vllm_engine064.model_dump(),
             "engine_dockerfile_config": {"VERSION":"v0.7.1"},
-            "default_cli_args": "--max_num_seq 256 --max_model_len 16000"
+            "default_cli_args": "--max_num_seq 256 --max_model_len 16000 --chat-template emd/models/chat_templates/deepseek_r1_distill.jinja"
 })
 
 vllm_deepseek_r1_distill_llama_engine071 = vllm_deepseek_r1_distill_qwen_engine071

src/emd/models/services.py

@@ -24,6 +24,25 @@
     support_cn_region = True
 )
 
+sagemaker_old_service = Service(
+    cfn_parameters={
+        "ECRImageURI":"ecr_repo_uri",
+        "InstanceType":"instance_type",
+        "ModelId": "model_id",
+        "ModelTag":"model_tag",
+        "FrameWorkType":"framework_type",
+        "ServiceType":"service_type",
+        "EngineType":"engine_type",
+        "Region":"region",
+        "MaxCapacity": ValueWithDefault(name="max_capacity",default=1),
+        "AutoScalingTargetValue": ValueWithDefault(name="auto_scaling_target_value",default=10)
+    },
+    name = "Amazon SageMaker AI Real-time inference",
+    service_type=ServiceType.SAGEMAKER_OLDER,
+    description="Amazon SageMaker Real-time inference provides low-latency, interactive inference through fully managed endpoints that support autoscaling. \n(https://docs.aws.amazon.com/sagemaker/latest/dg/realtime-endpoints.html)",
+    support_cn_region = True
+)
+
 sagemaker_async_service = Service(
     cfn_parameters={
         "ECRImageURI":"ecr_repo_uri",