ttnn.deallocate

ttnn.deallocate(tensor: ttnn.Tensor, force: bool | None = True) → `None`

Deallocates device tensor. Releases the resources for ttnn.Tensor tensor explicitly.

Parameters:

  • tensor (ttnn.Tensor) – the input tensor.

  • force (bool, optional) – force deallocation even if the buffer may have multiple references. Defaults to True.

Returns:

None – deallocates the tensor.

Example

>>> device_id = 0
>>> device = ttnn.open_device(device_id=device_id)
>>> tensor = ttnn.to_device(ttnn.from_torch(torch.randn((10, 64, 32), dtype=torch.bfloat16)), device)
>>> tensor = ttnn.to_layout(tensor, layout=ttnn.TILE_LAYOUT)
>>> ttnn.deallocate(tensor=tensor, force=False)