biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[MISC] Replace c10::optional with std::optional (#11730)

Browse Source 
Signed-off-by: Lu Fang <lufang@fb.com>
This commit is contained in:
Lu Fang
2025-01-04 17:20:34 -08:00
committed by GitHub
parent 47831430cc
commit 4068f4b5b5
24 changed files with 136 additions and 136 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
csrc/rocm/attention.cu
View File

@@ -928,7 +928,7 @@ void paged_attention_custom_launcher(
torch::Tensor& tmp_out, torch::Tensor& query, torch::Tensor& key_cache,
torch::Tensor& value_cache, const int num_kv_heads, float scale,
torch::Tensor& block_tables, torch::Tensor& context_lens,
int max_context_len, const c10::optional<torch::Tensor>& alibi_slopes,
int max_context_len, const std::optional<torch::Tensor>& alibi_slopes,
float k_scale, float v_scale) {
int num_seqs = query.size(0);
int num_heads = query.size(1);
@@ -1086,7 +1086,7 @@ void paged_attention(
torch::Tensor& block_tables, // [num_seqs, max_num_blocks_per_seq]
torch::Tensor& context_lens, // [num_seqs]
int64_t block_size, int64_t max_context_len,
const c10::optional<torch::Tensor>& alibi_slopes,
const std::optional<torch::Tensor>& alibi_slopes,
const std::string& kv_cache_dtype, double k_scale, double v_scale) {
const int head_size = query.size(2);
if (kv_cache_dtype == "auto") {