[Tensor] Support Read/Save Functionalities for Quantized Type

This PR enhances the existing functionality by supporting reading and saving operations on quantized type tensors.
This change can ensure that quantized tensors can accurately retrieve and store their data from binary files.

**Self-evaluation:**
1. Build test: [X]Passed [ ]Failed [ ]Skipped
2. Run test:   [X]Passed [ ]Failed [ ]Skipped

Signed-off-by: Donghyeon Jeong <[email protected]>

nntrainer/tensor/char_tensor.cpp

@@ -348,6 +348,8 @@ void CharTensor::print(std::ostream &out) const {
     }
     out.copyfmt(init);
   }
+
+  /// @todo print quantization information
 }
 
 void CharTensor::copy(const void *buf) {
@@ -364,4 +366,14 @@ void CharTensor::copy(const void *buf) {
   }
 }
 
+void CharTensor::save_quantization_info(std::ostream &file) {
+  checkedWrite(file, (char *)&axis, sizeof(uint8_t),
+               "[CharTensor::save] failed to write quantization information");
+}
+
+void CharTensor::read_quantization_info(std::ifstream &file) {
+  checkedRead(file, (char *)&axis, sizeof(uint8_t),
+              "[CharTensor::read] failed to read quantization information");
+}
+
 } // namespace nntrainer

nntrainer/tensor/char_tensor.h

@@ -217,7 +217,22 @@ class CharTensor : public TensorBase {
    */
   void print(std::ostream &out) const override;
 
+  /**
+   * @copydoc TensorBase::save_quantization_info()
+   */
+  void save_quantization_info(std::ostream &file) override;
+
+  /**
+   * @copydoc TensorBase::read_quantization_info()
+   */
+  void read_quantization_info(std::ifstream &file) override;
+
 private:
+  /**
+   * @brief quantization axis
+   */
+  uint8_t axis;
+
   /**
    * @brief copy a buffer to @a this, the caller has to ensure that @a this is
    * initialized otherwise undefined behavior

nntrainer/tensor/tensor.cpp

@@ -1036,9 +1036,12 @@ void Tensor::save(std::ostream &file) {
   NNTR_THROW_IF(!getContiguous(), std::invalid_argument)
     << getName() << " is not contiguous, cannot save.";
 
-  std::streamsize sz = static_cast<std::streamsize>(bytes());
+  /// @note Save quantization information which only works on Quantized Tensor
+  itensor->save_quantization_info(file);
+
+  std::streamsize sz = static_cast<std::streamsize>(bytes() + scale_size());
   NNTR_THROW_IF(sz < 0, std::invalid_argument)
-    << "save size: " << bytes()
+    << "save size: " << bytes() + scale_size()
     << " is too big. It cannot be represented by std::streamsize";
 
   checkedWrite(file, getData<char>(), sz, "[Tensor::save] operation failed");
@@ -1049,10 +1052,13 @@ void Tensor::read(std::ifstream &file) {
   NNTR_THROW_IF(!getContiguous(), std::invalid_argument)
     << getName() << " is not contiguous, cannot read.";
 
-  std::streamsize sz = static_cast<std::streamsize>(bytes());
+  /// @note Read quantization information which only works on Quantized Tensor
+  itensor->read_quantization_info(file);
+
+  std::streamsize sz = static_cast<std::streamsize>(bytes() + scale_size());
 
   NNTR_THROW_IF(sz < 0, std::invalid_argument)
-    << "read size: " << bytes()
+    << "read size: " << bytes() + scale_size()
     << " is too big. It cannot be represented by std::streamsize";
 
   checkedRead(file, getData<char>(), sz, "[Tensor::read] operation failed");

nntrainer/tensor/tensor_base.h

@@ -582,6 +582,16 @@ class TensorBase {
   size_t getIndex(unsigned int b, unsigned int c, unsigned int h,
                   unsigned int w) const noexcept;
 
+  /**
+   * @brief     Save quantization information
+   */
+  virtual void save_quantization_info(std::ostream &file) {}
+
+  /**
+   * @brief     Read quantization information
+   */
+  virtual void read_quantization_info(std::ifstream &file) {}
+
   /**
    * @brief     Get size of current tensor
    * @retval    unsigned int size of the current tensor