rms_norm

rms_norm[type: DType, rank: Int, input_0_fn: fn[Int, Int](IndexList[$1]) capturing -> SIMD[type, $0], /, target: StringSlice[StaticConstantOrigin] = __init__[__mlir_type.!kgen.string]("cpu"), multiply_before_cast: Bool = True](shape: IndexList[rank], gamma: NDBuffer[type, 1, origin], epsilon: SIMD[type, 1], weight_offset: SIMD[type, 1], output: NDBuffer[type, rank, origin], ctx: DeviceContextPtr)

Was this page helpful?

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

Thank you for helping us improve!