ttnn.clamp_bw

ttnn.clamp_bw = Operation(python_fully_qualified_name='ttnn.clamp_bw', function=<ttnn._ttnn.operations.unary_backward.clamp_bw_t object>, preprocess_golden_function_inputs=<function default_preprocess_golden_function_inputs>, golden_function=<function <lambda>>, postprocess_golden_function_outputs=<function default_postprocess_golden_function_outputs>, is_cpp_operation=True, is_experimental=False)

Performs backward operations for clamp on input_tensor, min, max with given grad_tensor.

Parameters:
Keyword Arguments:
  • min (float, optional) – Minimum value. Defaults to None.

  • max (float, optional) – Maximum value. Defaults to None.

  • memory_config (ttnn.MemoryConfig, optional) – memory configuration for the operation. Defaults to None.

Returns:

List of ttnn.Tensor – the output tensor.

Note

Supported dtypes, layouts, and ranks:

Dtypes

Layouts

Ranks

BFLOAT16

TILE

2, 3, 4

Only one of min or max value can be None.

Example

>>> grad_tensor = ttnn.from_torch(torch.tensor([[1, 2], [3, 4]], dtype=torch.bfloat16), layout=ttnn.TILE_LAYOUT, device=device)
>>> input = ttnn.from_torch(torch.tensor([[1, 2], [3, 4]], dtype=torch.bfloat16, requires_grad=True), layout=ttnn.TILE_LAYOUT, device=device)
>>> min = 0.5
>>> max = 2.0
>>> output = ttnn.clamp_bw(grad_tensor, input, min, max)