Question

我尝试重新训练开始！！当我使用tensorflow 1.2.0和1.2.0训练脚本时，我会出错：

 2019-01-21 22:02:38.904542: W
 tensorflow/core/framework/op_def_util.cc:355] Op
 BatchNormWithGlobalNormalization is deprecated. It will cease to work
 in GraphDef version 9. Use tf.nn.batch_normalization(). 2019-01-21
 22:02:39.036440: I tensorflow/core/platform/cpu_feature_guard.cc:141]
 Your CPU supports instructions that this TensorFlow binary was not
 compiled to use: AVX2 Exception in thread "main"
 java.lang.IllegalArgumentException: No Operation named [softmax] in
 the Graph at
 org.tensorflow.Session$Runner.operationByName(Session.java:372) at
 org.tensorflow.Session$Runner.parseOutput(Session.java:381)

当我使用tensorflow 1.12.0和1.12.0训练模型时，我会出错：

 Exception in thread "main" java.lang.IllegalArgumentException: Invalid  
 GraphDef   at org.tensorflow.Graph.importGraphDef(Native Method)   at  
 org.tensorflow.Graph.importGraphDef(Graph.java:130)

我现在不怎么训练模型？我使用InceptionV3 我的命令文件是：python D：//retrain.py --image_dir D：\ train 火车日志：https://nofile.io/f/QOCUJz8HW3h/log1.12.txt

Main java code : File file = new File("Java.jpg"); Image image = new
 Image(file.toURI().toString()); imga.setImage(image); String
 modelpath= "C:\\Users\\AngruAdminAlex\\Documents\\inception_dec_2015";
 float verison=(float) 1.0; System.out.println("Opening: " +
 modelpath); modelselected = true; graphDef =
 readAllBytesOrExit(Paths.get(modelpath,
 "tensorflow_inception_graph.pb")); labels =
 readAllLinesOrExit(Paths.get(modelpath,
 "imagenet_comp_graph_label_strings.txt")); File file1 = new File
 ("Java.jpg"); String imagepath="Java.jpg"; System.out.println("Image
 Path: " + imagepath); BufferedImage img = ImageIO.read(file1); //read
 image System.out.println("Reading complete."); byte[] imageBytes =
 readAllBytesOrExit(Paths.get(imagepath)); byte data[] = imageBytes;
 FileOutputStream out = new FileOutputStream("buffer.rawfile");
 out.write(data); out.close(); try (Tensor image1 =
 Tensor.create(imageBytes)) { float[] labelProbabilities =
 executeInceptionGraph(graphDef, image1); int bestLabelIdx =
 maxIndex(labelProbabilities); System.out.println( String.format( "BEST
 MATCH: %s (%.2f likely)", labels.get(bestLabelIdx),
 labelProbabilities[bestLabelIdx] * 100f)); results.setText(
 String.format( "BEST MATCH: %s (%.2f likely)",
 labels.get(bestLabelIdx), labelProbabilities[bestLabelIdx] * 100f)); }
 }

Answer 1

如何为此代码训练InceptionV3

在代码更改中
Tensor result = s.runner().feed("DecodeJpeg/contents",image).fetch("softmax").run().get(0)) { 到Tensor result = s.runner().feed("DecodeJpeg/contents", image).fetch("final_result").run().get(0)) {
从此处下载Tensorflw本机库：Here
第三步是解压缩.zip之类的.jar文件并将lib文件复制到您的项目文件夹中。
只需从以下网址获取retrain.py：here
打开命令提示符并输入命令： python (path to retrain.py) --image_dir (path to image) -- output_graph (path to graph files(.pb) -- output_labels(path to label(.txt)
查找.pb文件，将其重命名为label.pb并在代码中使用
享受编码

如何重新训练inceptionV3？

1 个答案: