Method from_blob

tinygrad/tensor.py:505–516

Exposes the pointer as a Tensor without taking ownership of the original data. The pointer must remain valid for the entire lifetime of the created Tensor. You can pass in `dtype` and `device` keyword arguments to control the data type and device of the tensor. Additionally, all ot

(ptr:int, shape:tuple[int, ...], **kwargs)

  @staticmethod
  def from_blob(ptr:int, shape:tuple[int, ...], **kwargs) -> Tensor:
    """
    Exposes the pointer as a Tensor without taking ownership of the original data.
    The pointer must remain valid for the entire lifetime of the created Tensor.

    You can pass in `dtype` and `device` keyword arguments to control the data type and device of the tensor.
    Additionally, all other keyword arguments are passed to the constructor of the tensor.
    """
    r = Tensor.empty(*shape, **kwargs)
    assert isinstance(r.device, str)
    cast(Buffer, r.uop.buffer).allocate(external_ptr=ptr)
    return r

  @staticmethod
  def from_url(url:str, gunzip:bool=False, **kwargs) -> Tensor: