[4/N] Expose HiSparse cache metrics via KV connector stats (#56061)

Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
Nicolò Lucchesi
2026-09-12 10:47:55 -07:00
committed by GitHub
co-authored by Lucas Wilkinson OpenAI Codex
parent dff76bc3e8
commit 29332cf936
9 changed files with 308 additions and 8 deletions
+7 -6
View File
@@ -999,16 +999,17 @@ STABLE_TORCH_LIBRARY_FRAGMENT(_C_cache_ops, ops) {
" Tensor? request_state_indices,"
" int region_stride,"
" Tensor(a!)? miss_mask=None,"
" Tensor(b!)? attention_indices=None,"
" Tensor(b!)? stats=None,"
" Tensor(c!)? attention_indices=None,"
" int attention_block_stride=0,"
" Tensor? request_ids=None,"
" Tensor? source_block_table=None,"
" int source_block_size=0,"
" Tensor(c!)? resolved_global_indices=None,"
" Tensor(d!)? valid_counts=None,"
" Tensor(e!)? swap_host_physical_rows=None,"
" Tensor(f!)? swap_device_physical_rows=None,"
" Tensor(g!)? swap_counts=None,"
" Tensor(d!)? resolved_global_indices=None,"
" Tensor(e!)? valid_counts=None,"
" Tensor(f!)? swap_host_physical_rows=None,"
" Tensor(g!)? swap_device_physical_rows=None,"
" Tensor(h!)? swap_counts=None,"
" Tensor? resident_block_table=None,"
" int resident_block_size=0,"
" int resident_null_block=0) -> ()");