update pa tc interfaces

csrc/ops.h

@@ -52,8 +52,8 @@ void paged_attention_v1_opt_tc(
     torch::Tensor& value_cache, int64_t num_kv_heads, double scale,
     torch::Tensor& block_tables, torch::Tensor& seq_lens, int64_t block_size,
     int64_t max_seq_len, const c10::optional<torch::Tensor>& alibi_slopes,
-    const std::string& kv_cache_dtype, double kv_scale, const int64_t tp_rank,
-    const int64_t blocksparse_local_blocks,
+    const std::string& kv_cache_dtype, double k_scale, double v_scale,
+    const int64_t tp_rank, const int64_t blocksparse_local_blocks,
     const int64_t blocksparse_vert_stride, const int64_t blocksparse_block_size,
     const int64_t blocksparse_head_sliding_step);
 
@@ -63,8 +63,8 @@ void paged_attention_v2_opt_tc(
     torch::Tensor& value_cache, int64_t num_kv_heads, double scale,
     torch::Tensor& block_tables, torch::Tensor& seq_lens, int64_t block_size,
     int64_t max_seq_len, const c10::optional<torch::Tensor>& alibi_slopes,
-    const std::string& kv_cache_dtype, double kv_scale, const int64_t tp_rank,
-    const int64_t blocksparse_local_blocks,
+    const std::string& kv_cache_dtype, double k_scale, double v_scale,
+    const int64_t tp_rank, const int64_t blocksparse_local_blocks,
     const int64_t blocksparse_vert_stride, const int64_t blocksparse_block_size,
     const int64_t blocksparse_head_sliding_step);
 

csrc/torch_bindings.cpp

@@ -76,20 +76,20 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
   ops.impl("paged_attention_v2_opt", torch::kCUDA, &paged_attention_v2_opt);
 
   // Compute the attention between an input query and the cached
-  // keys/values using PagedAttention. (opt_tc)
+  // keys/values using PagedAttention. (opt)
   ops.def(
       "paged_attention_v1_opt_tc("
       "    Tensor! out, Tensor query, Tensor key_cache,"
       "    Tensor value_cache, int num_kv_heads, float scale,"
       "    Tensor block_tables, Tensor seq_lens, int block_size,"
       "    int max_seq_len, Tensor? alibi_slopes,"
-      "    str kv_cache_dtype, float kv_scale, int tp_rank,"
-      "    int blocksparse_local_blocks,"
+      "    str kv_cache_dtype, float k_scale, float v_scale,"
+      "    int tp_rank, int blocksparse_local_blocks,"
       "    int blocksparse_vert_stride, int blocksparse_block_size,"
       "    int blocksparse_head_sliding_step) -> ()");
   ops.impl("paged_attention_v1_opt_tc", torch::kCUDA, &paged_attention_v1_opt_tc);
 
-  // PagedAttention V2 (opt_tc).
+  // PagedAttention V2 (opt). 
   ops.def(
       "paged_attention_v2_opt_tc("
       "    Tensor! out, Tensor exp_sums, Tensor max_logits,"
@@ -97,8 +97,8 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
       "    Tensor value_cache, int num_kv_heads, float scale,"
       "    Tensor block_tables, Tensor seq_lens, int block_size,"
       "    int max_seq_len, Tensor? alibi_slopes,"
-      "    str kv_cache_dtype, float kv_scale, int tp_rank,"
-      "    int blocksparse_local_blocks,"
+      "    str kv_cache_dtype, float k_scale, float v_scale,"
+      "    int tp_rank, int blocksparse_local_blocks,"
       "    int blocksparse_vert_stride, int blocksparse_block_size,"
       "    int blocksparse_head_sliding_step) -> ()");
   ops.impl("paged_attention_v2_opt_tc", torch::kCUDA, &paged_attention_v2_opt_tc);