For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Mojo struct
Struct_ep_combine_async
struct Struct_ep_combine_async
Registers the ep.combine_async graph op with the graph compiler.
Implemented traits
Methods
execute
static def execute[combine_dtype: DType, hidden_size: Int, top_k: Int, n_experts: Int, max_token_per_rank: Int, n_gpus_per_node: Int, n_nodes: Int, //](atomic_counters: T, input_tokens: T, src_info: T, send_ptrs: T, recv_ptrs: T, recv_count_ptrs: T, context: DeviceContext)
Execute the Expert Parallelism combine kernel.
Sends expert-processed output tokens back to their original devices asynchronously, without waiting for the transfers to complete.
Parameters:
- combine_dtype (
DType):DTypeused for the token payload during the combine phase (inferred). - hidden_size (
Int): Size of the model's hidden dimension (inferred). - top_k (
Int): Number of experts each token is routed to (inferred). - n_experts (
Int): Total number of experts across all GPUs (inferred). - max_token_per_rank (
Int): Maximum number of tokens per GPU (inferred). - n_gpus_per_node (
Int): Number of GPUs per node (inferred). - n_nodes (
Int): Number of physical nodes (inferred).
Args:
- atomic_counters (
T): Atomic counters coordinating work across thread blocks during the combine phase. - input_tokens (
T): Expert output tokens to send back to their original devices. Shape[num_tokens, hidden_size]. - src_info (
T): Source routing information from the dispatch phase recording the originating rank and token index for each token. Shape[num_tokens, 2]. - send_ptrs (
T): Send buffer pointers for the combine phase. - recv_ptrs (
T): Receive buffer pointers for the combine phase. - recv_count_ptrs (
T): Receive count buffer pointers tracking the number of tokens received per expert. - context (
DeviceContext): GPU device context for the current device.