ttnn.relu_max

ttnn.relu_max = Operation(python_fully_qualified_name='ttnn.relu_max', function=<ttnn._ttnn.operations.unary.relu_max_t object>, preprocess_golden_function_inputs=<function default_preprocess_golden_function_inputs>, golden_function=<function _golden_function_relu_max>, postprocess_golden_function_outputs=<function default_postprocess_golden_function_outputs>, is_cpp_operation=True, is_experimental=False)

Applies relu_max to input_tensor element-wise with upper_limit.

This function caps off the input to a max value and a min value of 0

\[\mathrm{output\_tensor}_i = \verb|relu_max|(\mathrm{input\_tensor}_i, \verb|upper_limit|)\]
Parameters:
  • input_tensor (ttnn.Tensor) – the input tensor.

  • upper_limit (float) – The max value for ReLU function.

Keyword Arguments:
  • memory_config (ttnn.MemoryConfig, optional) – Memory configuration for the operation. Defaults to None.

  • output_tensor (ttnn.Tensor, optional) – preallocated output tensor. Defaults to None.

  • queue_id (int, optional) – command queue id. Defaults to 0.

Returns:

ttnn.Tensor – the output tensor.

Note

Supported dtypes, layouts, and ranks:

Dtypes

Layouts

Ranks

BFLOAT16, BFLOAT8_B

TILE

2, 3, 4

System memory is not supported.

Example

>>> tensor = ttnn.from_torch(torch.tensor([[1, 2], [3, 4]], dtype=torch.bfloat16), dtype=ttnn.bfloat16, layout=ttnn.TILE_LAYOUT, device=device)
>>> upper_limit = 3
>>> output = ttnn.relu_max(tensor, upper_limit)