ByteBuffer outputBuffer = interpreter.getOutputTensor(0).asReadOnlyBuffer();
interpreter.run(input, null);
// Copy or read from outputBuffer (which may now be invalid).
[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["缺少我需要的資訊","missingTheInformationINeed","thumb-down"],["過於複雜/步驟過多","tooComplicatedTooManySteps","thumb-down"],["過時","outOfDate","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["示例/程式碼問題","samplesCodeIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-07-24 (世界標準時間)。"],[],[],null,["# Tensor\n\npublic interface **Tensor** \nA typed multi-dimensional array used in Tensorflow Lite.\n\nThe native handle of a `Tensor` is managed by `NativeInterpreterWrapper`, and does\nnot needed to be closed by the client. However, once the `NativeInterpreterWrapper` has\nbeen closed, the tensor handle will be invalidated.\n\n\u003cbr /\u003e\n\n### Nested Classes\n\n|-------|---|---|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| class | [Tensor.QuantizationParams](/edge/api/tflite/java/org/tensorflow/lite/Tensor.QuantizationParams) || Quantization parameters that corresponds to the table, `QuantizationParameters`, in the [TFLite Model schema file.](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/schema/schema.fbs) |\n\n### Public Methods\n\n|-----------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| abstract [ByteBuffer](https://developer.android.com/reference/java/nio/ByteBuffer.html) | [asReadOnlyBuffer](/edge/api/tflite/java/org/tensorflow/lite/Tensor#asReadOnlyBuffer())() Returns a read-only `ByteBuffer` view of the tensor data. |\n| abstract [DataType](/edge/api/tflite/java/org/tensorflow/lite/DataType) | [dataType](/edge/api/tflite/java/org/tensorflow/lite/Tensor#dataType())() Returns the [DataType](/edge/api/tflite/java/org/tensorflow/lite/DataType) of elements stored in the Tensor. |\n| abstract int | [numBytes](/edge/api/tflite/java/org/tensorflow/lite/Tensor#numBytes())() Returns the size, in bytes, of the tensor data. |\n| abstract int | [numDimensions](/edge/api/tflite/java/org/tensorflow/lite/Tensor#numDimensions())() Returns the number of dimensions (sometimes referred to as [rank](https://www.tensorflow.org/resources/dims_types.html#rank)) of the Tensor. |\n| abstract int | [numElements](/edge/api/tflite/java/org/tensorflow/lite/Tensor#numElements())() Returns the number of elements in a flattened (1-D) view of the tensor. |\n| abstract [Tensor.QuantizationParams](/edge/api/tflite/java/org/tensorflow/lite/Tensor.QuantizationParams) | [quantizationParams](/edge/api/tflite/java/org/tensorflow/lite/Tensor#quantizationParams())() Returns the quantization parameters of the tensor within the owning interpreter. |\n| abstract int\\[\\] | [shape](/edge/api/tflite/java/org/tensorflow/lite/Tensor#shape())() Returns the [shape](https://www.tensorflow.org/resources/dims_types.html#shape) of the Tensor, i.e., the sizes of each dimension. |\n| abstract int\\[\\] | [shapeSignature](/edge/api/tflite/java/org/tensorflow/lite/Tensor#shapeSignature())() Returns the original [shape](https://www.tensorflow.org/resources/dims_types.html#shape) of the Tensor, i.e., the sizes of each dimension - before any resizing was performed. |\n\nPublic Methods\n--------------\n\n#### public abstract [ByteBuffer](https://developer.android.com/reference/java/nio/ByteBuffer.html)\n**asReadOnlyBuffer**\n()\n\nReturns a read-only `ByteBuffer` view of the tensor data.\n\nIn general, this method is most useful for obtaining a read-only view of output tensor data,\n\\*after\\* inference has been executed (e.g., via [InterpreterApi.run(Object, Object)](/edge/api/tflite/java/org/tensorflow/lite/InterpreterApi#run(java.lang.Object,%20java.lang.Object))). In\nparticular, some graphs have dynamically shaped outputs, which can make feeding a predefined\noutput buffer to the interpreter awkward. Example usage:\n\n interpreter.run(input, null);\n ByteBuffer outputBuffer = interpreter.getOutputTensor(0).asReadOnlyBuffer();\n // Copy or read from outputBuffer.\n\nWARNING: If the tensor has not yet been allocated, e.g., before inference has been executed,\nthe result is undefined. Note that the underlying tensor pointer may also change when the\ntensor is invalidated in any way (e.g., if inference is executed, or the graph is resized), so\nit is \\*not\\* safe to hold a reference to the returned buffer beyond immediate use directly\nfollowing inference. Example \\*bad\\* usage:\n\n ByteBuffer outputBuffer = interpreter.getOutputTensor(0).asReadOnlyBuffer();\n interpreter.run(input, null);\n // Copy or read from outputBuffer (which may now be invalid).\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n##### Throws\n\n| [IllegalArgumentException](https://developer.android.com/reference/java/lang/IllegalArgumentException.html) | if the tensor data has not been allocated. |\n|-------------------------------------------------------------------------------------------------------------|--------------------------------------------|\n\n#### public abstract [DataType](/edge/api/tflite/java/org/tensorflow/lite/DataType)\n**dataType**\n()\n\nReturns the [DataType](/edge/api/tflite/java/org/tensorflow/lite/DataType) of elements stored in the Tensor. \n\n#### public abstract int\n**numBytes**\n()\n\nReturns the size, in bytes, of the tensor data. \n\n#### public abstract int\n**numDimensions**\n()\n\nReturns the number of dimensions (sometimes referred to as [rank](https://www.tensorflow.org/resources/dims_types.html#rank)) of the Tensor.\n\nWill be 0 for a scalar, 1 for a vector, 2 for a matrix, 3 for a 3-dimensional tensor etc.\n\n\u003cbr /\u003e\n\n#### public abstract int\n**numElements**\n()\n\nReturns the number of elements in a flattened (1-D) view of the tensor. \n\n#### public abstract [Tensor.QuantizationParams](/edge/api/tflite/java/org/tensorflow/lite/Tensor.QuantizationParams)\n**quantizationParams**\n()\n\nReturns the quantization parameters of the tensor within the owning interpreter.\n\nOnly quantized tensors have valid `QuantizationParameters`. For tensor that are not\nquantized, the values of scale and zero_point are both 0.\n\n\u003cbr /\u003e\n\n#### public abstract int\\[\\]\n**shape**\n()\n\nReturns the [shape](https://www.tensorflow.org/resources/dims_types.html#shape) of\nthe Tensor, i.e., the sizes of each dimension. \n\n##### Returns\n\n- an array where the i-th element is the size of the i-th dimension of the tensor. \n\n#### public abstract int\\[\\]\n**shapeSignature**\n()\n\nReturns the original [shape](https://www.tensorflow.org/resources/dims_types.html#shape) of the Tensor,\ni.e., the sizes of each dimension - before any resizing was performed. Unknown dimensions are\ndesignated with a value of -1. \n\n##### Returns\n\n- an array where the i-th element is the size of the i-th dimension of the tensor."]]