Mojo function

quantize_dynamic_scaled_fp4fp8

quantize_dynamic_scaled_fp4fp8[out_dtype: DType, scales_dtype: DType, in_dtype: DType, //, *, SF_VECTOR_SIZE: Int = 16, num_max_threads: Int = 512](ctx: DeviceContext, output_tile: TileTensor[out_dtype, output_tile.LayoutType, output_tile.origin, linear_idx_type=output_tile.linear_idx_type, element_size=output_tile.element_size], scales_tile: TileTensor[scales_dtype, scales_tile.LayoutType, scales_tile.origin, linear_idx_type=scales_tile.linear_idx_type, element_size=scales_tile.element_size], input_tile: TileTensor[in_dtype, input_tile.LayoutType, input_tile.origin, linear_idx_type=input_tile.linear_idx_type, element_size=input_tile.element_size], num_cols: Int, num_cols_padded: Int, tensor_sf: Float32 = 1)

Was this page helpful?

Thank you! We'll create more content like this.

Thank you for helping us improve!

😔 What went wrong?

Some code doesn’t work

It includes inaccurate information

It's missing information I need

It was difficult to understand

Other

Submit