[Kernel][Misc] Use TORCH_LIBRARY instead of PYBIND11_MODULE for custom ops (#5047)

This commit is contained in:
bnellnm
2024-06-09 16:23:30 -04:00
committed by GitHub
parent 5d7e3d0176
commit 5467ac3196
55 changed files with 833 additions and 451 deletions
+6 -6
View File
@@ -1,4 +1,4 @@
#include <torch/extension.h>
#include <torch/all.h>
#include <ATen/cuda/CUDAContext.h>
#include <c10/cuda/CUDAGuard.h>
@@ -127,7 +127,7 @@ void rotary_embedding(
// [num_tokens, num_heads * head_size]
torch::Tensor& key, // [batch_size, seq_len, num_kv_heads * head_size] or
// [num_tokens, num_kv_heads * head_size]
int head_size,
int64_t head_size,
torch::Tensor& cos_sin_cache, // [max_position, rot_dim]
bool is_neox) {
int64_t num_tokens = query.numel() / query.size(-1);
@@ -138,7 +138,7 @@ void rotary_embedding(
int64_t key_stride = key.stride(-2);
dim3 grid(num_tokens);
dim3 block(std::min(num_heads * rot_dim / 2, 512));
dim3 block(std::min<int64_t>(num_heads * rot_dim / 2, 512));
const at::cuda::OptionalCUDAGuard device_guard(device_of(query));
const cudaStream_t stream = at::cuda::getCurrentCUDAStream();
VLLM_DISPATCH_FLOATING_TYPES(query.scalar_type(), "rotary_embedding", [&] {
@@ -168,9 +168,9 @@ void batched_rotary_embedding(
// [num_tokens, num_heads * head_size]
torch::Tensor& key, // [batch_size, seq_len, num_kv_heads * head_size] or
// [num_tokens, num_kv_heads * head_size]
int head_size,
int64_t head_size,
torch::Tensor& cos_sin_cache, // [max_position, rot_dim]
bool is_neox, int rot_dim,
bool is_neox, int64_t rot_dim,
torch::Tensor& cos_sin_cache_offsets // [num_tokens]
) {
int64_t num_tokens = cos_sin_cache_offsets.size(0);
@@ -180,7 +180,7 @@ void batched_rotary_embedding(
int64_t key_stride = key.stride(-2);
dim3 grid(num_tokens);
dim3 block(std::min(num_heads * rot_dim / 2, 512));
dim3 block(std::min<int64_t>(num_heads * rot_dim / 2, 512));
const at::cuda::OptionalCUDAGuard device_guard(device_of(query));
const cudaStream_t stream = at::cuda::getCurrentCUDAStream();
VLLM_DISPATCH_FLOATING_TYPES(query.scalar_type(), "rotary_embedding", [&] {