For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).
Mojo struct
KVCacheCopyPagesD2H
struct KVCacheCopyPagesD2H
Registers the mo.kv_cache.copy_pages_d2h graph op with the graph compiler.
Implemented traits
Methods
execute
static def execute[dtype: DType, //, target: StringSlice[ImmStaticOrigin]](device_kv_blocks: ManagedTensorSlice[IOSpec[_, _].MutableInput, static_spec=device_kv_blocks.static_spec], host_kv_blocks: ManagedTensorSlice[IOSpec[_, _].MutableInput, static_spec=host_kv_blocks.static_spec], src_page_ids: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=src_page_ids.static_spec], dst_page_ids: ManagedTensorSlice[IOSpec[_, _].Input, static_spec=dst_page_ids.static_spec], layer_idx: UInt32, ctx: DeviceContext)
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!