ttnn.sharded_to_interleaved

ttnn.sharded_to_interleaved(input_tensor: ttnn.Tensor, memory_config: ttnn.MemoryConfig, *, output_dtype: ttnn.DataType | None) → None

Converts a tensor from sharded to interleaved memory layout

Parameters:

• input_tensor (ttnn.Tensor) – input tensor

• memory_config (ttnn.MemoryConfig) – Memory configuration for the operation, must be Interleaved.

Keyword Arguments:

output_dtype (Optional[ttnn.DataType]) – Output data type, defaults to same as input.

Example

>>> interleaved_tensor = ttnn.sharded_to_interleaved(tensor, ttnn.DRAM_MEMORY_CONFIG)