[XPU][CI] Make test_mamba_prefix_cache block-size agnostic (#58269)
Signed-off-by: Lin, Fanli <fanli.lin@intel.com> Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
@@ -464,13 +464,16 @@ class XPUPlatform(Platform):
|
||||
)
|
||||
cache_config.block_size = new_block_size
|
||||
logger.info(
|
||||
"[XPU]Setting attention block size to %d tokens to ensure multiple of %d, "
|
||||
"set mamba_page_size_padded to %d bytes accordingly, before was %d bytes.",
|
||||
"[XPU]Setting attention block size to %d tokens to ensure multiple of %d.",
|
||||
new_block_size,
|
||||
kernel_block_size,
|
||||
cache_config.mamba_page_size_padded,
|
||||
original_mamba_page_size_padded,
|
||||
)
|
||||
if original_mamba_page_size_padded is not None:
|
||||
logger.info(
|
||||
"[XPU]Scaled mamba_page_size_padded from %d to %d bytes accordingly.",
|
||||
original_mamba_page_size_padded,
|
||||
cache_config.mamba_page_size_padded,
|
||||
)
|
||||
|
||||
@classmethod
|
||||
def support_hybrid_kv_cache(cls) -> bool:
|
||||
|
||||
Reference in New Issue
Block a user