出现不可能的错误：grpc状态=状态代码。NOT_FOUND，details=“未找到请求的服务”

Question

我正在尝试提供keras.applications模型。这不是我第一次在tensorflow-server docker映像的帮助下这样做，但是当我尝试使用新发布的模型Nasnet Mobile时，我不知道为什么我的代码停止工作。我得到的错误是：

_会合地点：

我用标准程序导出模型

from keras import backend as K

K.set_learning_phase(0) # Deactivate train-only-layers like: batch norm and dropout
print(model.input)
print(model.output)

from tensorflow.python.saved_model import builder as saved_model_builder

export_path = 'export/nasnet/1' # should always end on int (model versioning)
builder = saved_model_builder.SavedModelBuilder(export_path)


from tensorflow.python.saved_model import tag_constants, signature_constants
from tensorflow.python.`enter code here`saved_model.signature_def_utils_impl import predict_signature_def
#from tensorflow.python.saved_model.signature_def_utils_impl import build_signature_def

in_tensors = dict()
out_tensors = dict()

sess =  K.get_session()

in_tensors['input'] = sess.graph.get_tensor_by_name('input_1:0')
out_tensors['predictions'] = sess.graph.get_tensor_by_name('predictions/Softmax:0')

prediction_signature = predict_signature_def(inputs=in_tensors,
                                            outputs=out_tensors)

# export the protobuf and its signatures
builder.add_meta_graph_and_variables(sess=sess,
                                    tags=[tag_constants.SERVING],
                                    signature_def_map={
        signature_constants.DEFAULT_SERVING_SIGNATURE_DEF_KEY:prediction_signature})

builder.save()

正常输出显示:

Tensor("input_1:0", shape=(?, 224, 224, 3), dtype=float32)
Tensor("predictions/Softmax:0", shape=(?, 1000), dtype=float32)


INFO:tensorflow:No assets to save.
INFO:tensorflow:No assets to write.
INFO:tensorflow:SavedModel written to: export/nasnet/1/saved_model.pb

b'export/nasnet/1/saved_model.pb'

根据grpc文档，这种错误甚至不应该由grpc产生。很自然，我将模型装载到docker容器中:

docker run -d -p 8500:8500 \
    --mount type=bind,source=$(pwd)/export/nasnet/,target=/models/nasnet \
    -e MODEL_NAME=nasnet -e TF_CPP_MIN_VLOG_LEVEL=0 \
    --name nasnet_tfserving \
    -t tensorflow/serving:1.10.0 

容器中只装载了一个模型，版本为1。是什么导致了这种状态代码？docker日志看起来正常:

Docker日志nasnet_tfserving2018-12-10 23:12:06.902367：我tensorflow_serving/model_servers/main.cc:157]构建单个TensorFlow模型文件配置：model_name： nasnetmodel_base_path： /models/nasnet2018-12-10 23:12:06.904872：我tensorflow_serving/model_servers/server_core.cc:462]添加/更新模型。2018-12-10 23:12:06.904932：我tensorflow_serving/model_servers/server_core.cc:517]（重新）添加模型： nasnet 2018-12-10 23:12:07.006261：我tensorflow_serving/core/basic_manager.cc:739]成功保留资源以加载servable{name： nasnet version： 1} 2018-12-10 23:12:07.006484： Itensorflow_serving/core/loader_harness.cc:66]批准加载servable version{name： nasnet version： 1} 2018-12-10 23:12:07.006539：我tensorflow_serving/core/loader_harness.cc:74]加载servable version{name： nasnet version： 1} 2018-12-10 23:12:07.006621：我外部/org_tensorflow/tenorflow/contrib/session_bundle/bundle_shim.cc:360]尝试在bundle-shim中加载本地SavedModelBundle来自： /models/nasnet/12018-12-10 23:12:07.006810： I外部/org_tensorflow/tenorflow/cc/saved_model/reader.cc:31]从： /models/nasnet/12018-12-10 23:12:07.257410： I外部/org_tensorflow/tenorflow/cc/saved_model/reader.cc:54]使用标签{service}读取元图2018-12-10 23:12:07.557033： I外部/org_tensorflow/tenorflow/core/平台/cpu_feature_guard.cc:141]您的CPU支持此TensorFlow二进制文件未编译为使用的指令：FMA 2018-12-10 23:12:08.726888： I外部/org_tensorflow/tenorflow/cc/saved_model/loader.cc:113]恢复SavedModel捆绑包。2018-12-10 23:12:09.814493： I外部/org_tensorflow/tenorflow/cc/saved_model/loader.cc:148]在SavedModel捆绑包上运行LegacyInitOp。2018-12-10 23:12:09.814631： I外部/org_tensorflow/tenorflow/cc/saved_model/loader.cc:233]标签{service}的SavedModel负载；状态：成功。花了2807935微秒。2018-12-10 23:12:09.814818：我tensorflow_serving/servables/tenorflow/saved_model_warmup.cc:83]在 /models/nasnet/1/assets.extra/tf_serving_warmup_requests找不到预热数据文件2018-12-10 23:12:09.815253：我tensorflow_serving/core/loader_harness.cc:86]成功加载可服务版本{name： nasnet version： 1} 2018-12-10 23:12:09.823923：我tensorflow_serving/model_servers/main.cc:327]在0.0.0.0:8500运行ModelServer...[警告]getaddrinfo：不支持节点名的地址系列[evhttp_server.cc： 235]RAW：进入事件循环...2018-12-10 23:12:09.832156：我tensorflow_serving/model_servers/main.cc:337]在：localhost:8501...

有什么意见吗？？

Answer 1

您的客户端代码似乎有问题。错误显示“未为请求找到可服务：最新（nasnest）”。但您使用“nasnet”作为模型名启动了tf服务器。只需在客户端代码中将“nasnest”更改为“nasnet”。