Skip to content

GitLab

Commit bbf9488b authored by zhuwenwen's avatar zhuwenwen
Browse files

change pa v1 to 128

parent 8ee4ae1f
Hide whitespace changes
Inline Side-by-side
Showing with 1 addition and 1 deletion
csrc/attention/attention_kernels.cu
View file @ bbf9488b
...@@ -689,7 +689,7 @@ __global__ void paged_attention_v2_reduce_kernel( ...@@ -689,7 +689,7 @@ __global__ void paged_attention_v2_reduce_kernel(
// TODO(woosuk): Tune NUM_THREADS. // TODO(woosuk): Tune NUM_THREADS.
template <typename T, typename CACHE_T, int BLOCK_SIZE, template <typename T, typename CACHE_T, int BLOCK_SIZE,
vllm::Fp8KVCacheDataType KV_DTYPE, bool IS_BLOCK_SPARSE, vllm::Fp8KVCacheDataType KV_DTYPE, bool IS_BLOCK_SPARSE,
int NUM_THREADS = 256> int NUM_THREADS = 128>
void paged_attention_v1_launcher( void paged_attention_v1_launcher(
torch::Tensor& out, torch::Tensor& query, torch::Tensor& key_cache, torch::Tensor& out, torch::Tensor& query, torch::Tensor& key_cache,
torch::Tensor& value_cache, int num_kv_heads, float scale, torch::Tensor& value_cache, int num_kv_heads, float scale,
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment