For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Tensor

#include "max/c/tensor.h"

Functions:

M_newTensorSpec()

M_TensorSpec *M_newTensorSpec(const int64_t *shape, int64_t rankSize, M_Dtype dtype, const char *tensorName, const M_Device *device)

Creates a tensor specification.

You need this in order to set the input tensors with M_borrowTensorInto().

When storing tensor data in memory, we always use a diminishing stride size. That is, earlier dimensions in the shape have larger strides than later dimensions. For example, a C array declared as int arr[1][2][3] would have a shape specified as {1, 2, 3}.

• Parameters:

  • shape – The shape of the tensor.
  • rankSize – The rank size of the tensor.
  • dtype – The datatype for the tensor.
  • tensorName – The name for the tensor. This string gets copied as part of the operation of M_newTensorSpec, so your original string need not remain valid after the completion of this call.
  • device – The device on which the tensor resides.

• Returns:

  A pointer to the tensor spec. You are responsible for the memory associated with the pointer returned. The memory can be deallocated by calling M_freeTensorSpec().

M_isDynamicRanked()

int M_isDynamicRanked(const M_TensorSpec *spec)

Returns if the given spec has a dynamic rank.

• Parameters:

  spec – The tensor spec.

• Returns:

  1 if the rank is dynamic. 0 otherwise.

M_getDimAt()

int64_t M_getDimAt(const M_TensorSpec *spec, size_t axis)

Gets the element at a particular axis.

• Parameters:

  • spec – The tensor spec.
  • axis – The requested axis

• Returns:

  The dimension at requested axis if the spec and axis are valid and has static rank. Otherwise, 0. A dimension equaling kDynamicDimensionValue indicates dynamic dimension e.g. batch-size of a model expecting a batched tensor.

M_getRank()

int64_t M_getRank(const M_TensorSpec *spec)

Gets the rank from the tensor spec.

• Parameters:

  spec – The tensor spec.

• Returns:

  The number of dimensions in the tensor spec if the spec is static and valid, kDynamicRankValue if dynamic. Otherwise, 0.

M_getDtype()

M_Dtype M_getDtype(const M_TensorSpec *spec)

Gets the datatype from the tensor spec.

• Parameters:

  spec – The tensor spec.

• Returns:

  The element type from the tensor spec if the tensor spec is valid. Otherwise, M_UNKNOWN.

M_getName()

const char *M_getName(M_TensorSpec *spec)

Gets the name of the tensor from the tensor spec.

• Parameters:

  spec – The tensor spec.

• Returns:

  A null-terminated string containing the name of the tensor if the spec is valid. Otherwise, NULL. The memory associated with the returned string is owned by spec.

M_newAsyncTensorMap()

M_AsyncTensorMap *M_newAsyncTensorMap(const M_RuntimeContext *context)

Creates a map of tensor names to async tensors.

• Parameters:

  context – The runtime context.

• Returns:

  A pointer to the tensor map. You are responsible for the memory associated with the pointer returned. The memory can be deallocated by calling M_freeAsyncTensorMap().

M_borrowTensorInto()

void M_borrowTensorInto(M_AsyncTensorMap *tensors, void *input, const M_TensorSpec *tensorSpec, M_Status *status)

Adds a tensor to the tensor map.

You are responsible for the lifetime of the input tensor data. Its data gets “borrowed” into the Tensor Map.

• Parameters:

  • tensors – The tensor map, from M_newAsyncTensorMap().
  • input – The input tensor data.
  • tensorSpec – The tensor spec, from M_newTensorSpec(). This gets copied as part of the operation of M_borrowTensorInto, so your original tensorSpec need not exist through the lifetime of the tensor map.
  • status – The status object for reporting errors.

M_getTensorByNameFrom()

M_AsyncTensor *M_getTensorByNameFrom(M_AsyncTensorMap *tensorMap, const char *name, M_Status *status)

Gets a tensor from the tensor map by name.

• Parameters:

  • tensorMap – The tensor map.
  • name – The name of the tensor.
  • status – The status object for reporting errors.

• Returns:

  A pointer to the tensor. You are responsible for the memory associated with the pointer returned. The memory can be deallocated by calling M_freeTensor(). The held tensor inside the return value is simply borrowed from the corresponding input M_AsyncTensorMap. If the tensor map or name are invalid, a NULL pointer is returned and the status parameter contains an error message.

M_getTensorNumElements()

size_t M_getTensorNumElements(const M_AsyncTensor *tensor)

Gets the number of elements for the tensor.

• Parameters:

  tensor – The tensor which must not be NULL.

• Returns:

  The number of elements for the given tensor.

M_getTensorType()

M_Dtype M_getTensorType(const M_AsyncTensor *tensor)

Gets the corresponding M_Dtype for the tensor.

• Parameters:

  tensor – The tensor which must not be NULL.

• Returns:

  The corresponding M_Dtype for the tensor.

M_getTensorData()

const void *M_getTensorData(const M_AsyncTensor *tensor)

Gets a pointer to underlying data of the tensor.

• Parameters:

  tensor – The tensor which must not be NULL.

• Returns:

  A pointer to the underlying data of the tensor. This pointer is valid for the lifetime of the underlying tensor.

M_getTensorSpec()

M_TensorSpec *M_getTensorSpec(const M_AsyncTensor *tensor)

Gets a Tensor Spec for the tensor.

• Parameters:

  tensor – The tensor.

• Returns:

  The tensor spec for the tensor if the tensor is valid. Otherwise, NULL.

M_getDeviceTypeFromSpec()

M_DeviceType M_getDeviceTypeFromSpec(const M_TensorSpec *spec)

Gets the device type from a tensor specification.

• Parameters:

  spec – The tensor spec.

• Returns:

  The device type (CPU or GPU).

M_getDeviceIdFromSpec()

int M_getDeviceIdFromSpec(const M_TensorSpec *spec)

Gets the device ID from a tensor specification.

• Parameters:

  spec – The tensor spec.

• Returns:

  The device ID. Returns 0 if the spec is invalid.

M_getTensorDevice()

M_Device *M_getTensorDevice(const M_AsyncTensor *tensor)

Gets the device on which a tensor resides.

• Parameters:

  tensor – The tensor.

• Returns:

  The device on which the tensor resides, or NULL if the tensor is invalid. The caller owns the returned device and must free it with M_freeDevice().

M_copyTensorToDevice()

M_AsyncTensor *M_copyTensorToDevice(M_AsyncTensor *tensor, M_Device *device, M_Status *status)

Copies a tensor to a different device.

Creates a copy of the tensor on the specified device.

• Parameters:

  • tensor – The tensor to copy.
  • device – The target device.
  • status – The status object for reporting errors.

• Returns:

  A pointer to the tensor on the target device. The caller owns the returned memory and must deallocate it by calling M_freeTensor(). Returns NULL if the operation fails, with an error message in the status.

M_freeTensor()

void M_freeTensor(M_AsyncTensor *tensor)

Deallocates the memory for the tensor. No-op if tensor is NULL.

• Parameters:

  tensor – The tensor to deallocate.

M_freeTensorNameArray()

void M_freeTensorNameArray(M_TensorNameArray *names)

Deallocates the memory for the array of tensor names. No-op if names is NULL.

• Parameters:

  names – The tensor names to deallocate.

M_freeTensorSpec()

void M_freeTensorSpec(M_TensorSpec *spec)

Deallocates the memory for the tensor spec. No-op if spec is NULL.

• Parameters:

  spec – The tensor spec to deallocate.

M_freeAsyncTensorMap()

void M_freeAsyncTensorMap(M_AsyncTensorMap *tensorMap)

Deallocates the memory for the tensor map. No-op if tensorMap is NULL.

• Parameters:

  tensorMap – The tensor map to deallocate.

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit