Misha GoinandClaude
d79b52b182
Signed-off-by: mgoin <mgoin64@gmail.com> Co-authored-by: Claude <noreply@anthropic.com>
25 lines
1.0 KiB
C
25 lines
1.0 KiB
C
/*
|
|
* Adapted from
|
|
* https://github.com/pytorch/pytorch/blob/v2.0.1/aten/src/ATen/Dispatch.h
|
|
*/
|
|
#pragma once
|
|
|
|
#include <torch/all.h>
|
|
|
|
// Need a special dispatch case macro since we will nest the FP8 dispatch.
|
|
// Instead of the usual 'scalar_t', this names the dispatched type 'fp8_t'.
|
|
#define AT_DISPATCH_FP8_CASE(enum_type, ...) \
|
|
AT_PRIVATE_CASE_TYPE_USING_HINT(enum_type, fp8_t, __VA_ARGS__)
|
|
|
|
// ROCm devices might use either fn or fnuz, so set up dispatch table for both.
|
|
// A host-based check at runtime will create a preferred FP8 type for ROCm
|
|
// such that the correct kernel is dispatched.
|
|
#define VLLM_DISPATCH_CASE_FP8_TYPES(...) \
|
|
AT_DISPATCH_FP8_CASE(at::ScalarType::Float8_e4m3fn, __VA_ARGS__) \
|
|
AT_DISPATCH_FP8_CASE(at::ScalarType::Float8_e4m3fnuz, __VA_ARGS__)
|
|
|
|
// When using this dispatch macro, the type is 'fp8_t' not 'scalar_t'.
|
|
// See AT_DISPATCH_FP8_CASE above.
|
|
#define VLLM_DISPATCH_FP8_TYPES(TYPE, NAME, ...) \
|
|
AT_DISPATCH_SWITCH(TYPE, NAME, VLLM_DISPATCH_CASE_FP8_TYPES(__VA_ARGS__))
|