ttnn.softplus_bw
- ttnn.softplus_bw = Operation(python_fully_qualified_name='ttnn.softplus_bw', function=<ttnn._ttnn.operations.unary_backward.softplus_bw_t object>, preprocess_golden_function_inputs=<function default_preprocess_golden_function_inputs>, golden_function=<function _golden_function>, postprocess_golden_function_outputs=<function default_postprocess_golden_function_outputs>, is_cpp_operation=True, is_experimental=False)
-
Performs backward operations for softplus on
input_tensor
,beta
,threshold
with givengrad_tensor
.- Parameters:
-
grad_tensor (ttnn.Tensor) – the input gradient tensor.
input_tensor (ttnn.Tensor) – the input tensor.
- Keyword Arguments:
-
beta (float, optional) – Beta value for the Softplus formula . Defaults to 1.
threshold (float, optional) – Threshold value. Defaults to 20.
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
For more details about BFLOAT8_B, refer to the BFLOAT8_B limitations.
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) >>> output = ttnn.softplus_bw(grad_tensor, input, beta = 1, threshold = 20