reshape¶

nvtripy.reshape(input: Tensor, shape: Sequence[int | DimensionSize]) → Tensor[source]¶

Returns a new tensor with the contents of the input tensor in the specified shape.

Parameters:

input (Tensor) – The input tensor.
shape (Sequence[int | DimensionSize]) – The desired compatible shape. If a shape dimension is -1, its value is inferred based on the other dimensions and the number of elements in the input. Atmost one dimension can be -1.

Returns:

A new tensor with the specified shape.

Return type:

Tensor

INPUT REQUIREMENTS:: input.dtype is one of [float32, float16, bfloat16, int4, int8, int32, int64, bool]
OUTPUT GUARANTEES:: return[0].dtype == input.dtype

Example

input = tp.iota((2, 3), dtype=tp.float32)
output = tp.reshape(input, (1, 6))

Local Variables¶

>>> input
tensor(
    [[0, 0, 0],
     [1, 1, 1]], 
    dtype=float32, loc=gpu:0, shape=(2, 3))

>>> output
tensor(
    [[0, 0, 0, 1, 1, 1]], 
    dtype=float32, loc=gpu:0, shape=(1, 6))