Fix Bug 'get_cpu_copy not Implemented' in pd offloading mode (#7982)
This commit is contained in:
@@ -526,6 +526,12 @@ class PagedTokenToKVPoolAllocator(BaseTokenToKVPoolAllocator):
|
|||||||
self.is_not_in_free_group = True
|
self.is_not_in_free_group = True
|
||||||
self.free_group = []
|
self.free_group = []
|
||||||
|
|
||||||
|
def get_cpu_copy(self, indices):
|
||||||
|
return self._kvcache.get_cpu_copy(indices)
|
||||||
|
|
||||||
|
def load_cpu_copy(self, kv_cache_cpu, indices):
|
||||||
|
return self._kvcache.load_cpu_copy(kv_cache_cpu, indices)
|
||||||
|
|
||||||
|
|
||||||
def alloc_extend_kernel_ascend(
|
def alloc_extend_kernel_ascend(
|
||||||
prefix_lens,
|
prefix_lens,
|
||||||
|
|||||||
Reference in New Issue
Block a user