docker run -p 8501:8501 --mount type=bind,source=/xxx/traffic_prediction_gru,target=/models/traffic_prediction_gru -e MODEL_NAME=traffic_prediction_gru -t tensorflow/serving
若遇到错误提示
Could not find base path /models/model for servable model
可能是因为model_base_path没有配置,可在上述指令后加上一下语句
docker run -p 8501:8501 --mount type=bind,source=/xxx/traffic_prediction_gru,target=/models/traffic_prediction_gru -e MODEL_NAME=traffic_prediction_gru -t tensorflow/serving --model_base_path=/models/half_plus_two