[6/N][KV-Cache Layout Refactor] Standardize KV cache layout (#51718)

Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
This commit is contained in:
Lucas Wilkinson
2026-08-21 17:47:49 -07:00
committed by GitHub
parent 9ff7041b55
commit 8bdc70ec7b
150 changed files with 4006 additions and 5815 deletions
-3
View File
@@ -618,9 +618,7 @@ class GPUModelRunner(LoRAModelRunnerMixin):
self.kv_caches,
self.compilation_config.static_forward_context,
self.kv_cache_config,
self.attn_groups,
self.device,
self.cache_config.cache_dtype,
self.kernel_block_sizes,
self.vllm_config,
)
@@ -632,7 +630,6 @@ class GPUModelRunner(LoRAModelRunnerMixin):
self.device,
attn_groups_iter=(g for groups in self.attn_groups for g in groups),
kernel_block_sizes=self.kernel_block_sizes,
cache_dtype=self.cache_config.cache_dtype,
static_forward_context=self.compilation_config.static_forward_context,
)