Curl 查询 tensorflow 服务 model 以预测 API 中断

Question

I'm using this tutorial TensorFlow Serving with Docker to query

curl -d '{"instances": [1.0, 2.0, 5.0]}' \ -X POST http://localhost:8501/v1/models/half_plus_two:predict

It returns

`C:\WINDOWS\system32>curl -d '{"instances": [1.0, 2.0, 5.0]}' -X POST http://localhost:8501/v1/models/half_plus_two:predict

curl: (3) [globbing] bad range in column 2
curl: (6) Could not resolve host: 2.0,
curl: (3) [globbing] unmatched close brace/bracket in column 4
{ "error": "JSON Parse error: Invalid value. at offset: 0" }`

But the docker is running fine.

    PS E:\git_portable> docker run -t --rm -p 8501:8501 -v "E:\git_portable\serving\tensorflow_serving\servables\tensorflow\testdata\saved_model_half_plus_two_cpu:/models/half_plus_two" -e MODEL_NAME=half_plus_two tensorflow/serving
2019-11-10 07:11:17.037045: I tensorflow_serving/model_servers/server.cc:85] Building single TensorFlow model file config:  model_name: half_plus_two model_base_path: /models/half_plus_two
2019-11-10 07:11:17.037797: I tensorflow_serving/model_servers/server_core.cc:462] Adding/updating models.
2019-11-10 07:11:17.037861: I tensorflow_serving/model_servers/server_core.cc:573]  (Re-)adding model: half_plus_two
2019-11-10 07:11:17.158245: I tensorflow_serving/core/basic_manager.cc:739] Successfully reserved resources to load servable {name: half_plus_two version: 123}
2019-11-10 07:11:17.158435: I tensorflow_serving/core/loader_harness.cc:66] Approving load for servable version {name: half_plus_two version: 123}
2019-11-10 07:11:17.158496: I tensorflow_serving/core/loader_harness.cc:74] Loading servable version {name: half_plus_two version: 123}
2019-11-10 07:11:17.158573: I external/org_tensorflow/tensorflow/cc/saved_model/reader.cc:31] Reading SavedModel from: /models/half_plus_two/00000123
2019-11-10 07:11:17.170610: I external/org_tensorflow/tensorflow/cc/saved_model/reader.cc:54] Reading meta graph with tags { serve }
2019-11-10 07:11:17.172642: I external/org_tensorflow/tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
2019-11-10 07:11:17.212202: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:202] Restoring SavedModel bundle.
2019-11-10 07:11:17.230431: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:151] Running initialization op on SavedModel bundle at path: /models/half_plus_two/00000123
2019-11-10 07:11:17.236016: I external/org_tensorflow/tensorflow/cc/saved_model/loader.cc:311] SavedModel load for tags { serve }; Status: success. Took 77445 microseconds.
2019-11-10 07:11:17.237262: I tensorflow_serving/servables/tensorflow/saved_model_warmup.cc:105] No warmup data file found at /models/half_plus_two/00000123/assets.extra/tf_serving_warmup_requests
2019-11-10 07:11:17.247605: I tensorflow_serving/core/loader_harness.cc:87] Successfully loaded servable version {name: half_plus_two version: 123}
2019-11-10 07:11:17.250931: I tensorflow_serving/model_servers/server.cc:353] Running gRPC ModelServer at 0.0.0.0:8500 ...
[warn] getaddrinfo: address family for nodename not supported
2019-11-10 07:11:17.252948: I tensorflow_serving/model_servers/server.cc:373] Exporting HTTP/REST API at:localhost:8501 ...

When I run plain curl to localhost, it returns fine.

C:\WINDOWS\system32>curl http://localhost:8501/v1/models/half_plus_two
{
 "model_version_status": [
  {
   "version": "123",
   "state": "AVAILABLE",
   "status": {
    "error_code": "OK",
    "error_message": ""
   }
  }
 ]
}

What am I doing wrong here?

Answer 1

We had the same issue. Here..

Windows's cmd doesn't support strings with single quotes. Use " and escape the inner ones with \".

in this link: Windows: curl with json data on the command line

now,

`C:\WINDOWS\system32>curl -d "{\"instances\": [1.0, 2.0, 5.0]}" \
-X POST http://127.0.0.1:8501/v1/models/half_plus_two:predict

{
    "predictions": [2.5, 3.0, 4.5
    ]
}`