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
DistributedEPCombine
struct DistributedEPCombine
Registers the mo.distributed.ep.combine graph op with the graph compiler.
Implemented traits
Methods
execute
static def execute[combine_dtype: DType, router_weights_dtype: DType, hidden_size: Int, top_k: Int, n_experts: Int, max_token_per_rank: Int, n_gpus_per_node: Int, n_nodes: Int, fused_shared_expert: Bool, has_epilogue_fusion: Bool, //, target: StringSlice[ImmStaticOrigin], _trace_name: StringSlice[ImmStaticOrigin]](output_tokens: _FusedOutputVariadicTensors[static_specs=output_tokens.static_specs], input_tokens: VariadicTensors[IOSpec[_, _].Input, static_specs=input_tokens.static_specs], src_info: VariadicTensors[IOSpec[_, _].Input, static_specs=src_info.static_specs], send_ptrs: VariadicTensors[IOSpec[_, _].Input, static_specs=send_ptrs.static_specs], recv_ptrs: VariadicTensors[IOSpec[_, _].Input, static_specs=recv_ptrs.static_specs], recv_count_ptrs: VariadicTensors[IOSpec[_, _].Input, static_specs=recv_count_ptrs.static_specs], router_weights: VariadicTensors[IOSpec[_, _].Input, static_specs=router_weights.static_specs], atomic_counters: VariadicTensors[IOSpec[_, _].MutableInput, static_specs=atomic_counters.static_specs], dev_ctxs: DeviceContextArray)
Multi-device fused Expert Parallelism combine with output fusion.
Parameters:
- combine_dtype (
DType):DTypeof the combined expert output tokens. - router_weights_dtype (
DType):DTypeof the router weights. - hidden_size (
Int): Size of the model's hidden dimension. - top_k (
Int): Number of experts each token is routed to. - n_experts (
Int): Total number of experts across all GPUs. - max_token_per_rank (
Int): Maximum number of tokens per GPU. - n_gpus_per_node (
Int): Number of GPUs per node. - n_nodes (
Int): Number of physical nodes. - fused_shared_expert (
Bool): Whether a shared expert is fused into the combine kernel. - has_epilogue_fusion (
Bool): Whether the combine output is fused with a downstream elementwise epilogue. - target (
StringSlice[ImmStaticOrigin]): Compile-time device target. - _trace_name (
StringSlice[ImmStaticOrigin]): Trace label for this op.
Args:
- output_tokens (
_FusedOutputVariadicTensors[static_specs=output_tokens.static_specs]): Fused output variadic tensors storing the combined expert outputs, one per device. - input_tokens (
VariadicTensors[IOSpec[_, _].Input, static_specs=input_tokens.static_specs]): Input variadic tensors of expert-processed tokens to combine, one per device. - src_info (
VariadicTensors[IOSpec[_, _].Input, static_specs=src_info.static_specs]): Source info tensors recording the originating rank and token index for each received token, one per device. - send_ptrs (
VariadicTensors[IOSpec[_, _].Input, static_specs=send_ptrs.static_specs]): Send buffer pointers for the combine phase, one per device. - recv_ptrs (
VariadicTensors[IOSpec[_, _].Input, static_specs=recv_ptrs.static_specs]): Receive buffer pointers for the combine phase, one per device. - recv_count_ptrs (
VariadicTensors[IOSpec[_, _].Input, static_specs=recv_count_ptrs.static_specs]): Receive count buffer pointers tracking tokens received per expert, one per device. - router_weights (
VariadicTensors[IOSpec[_, _].Input, static_specs=router_weights.static_specs]): Router weights used to scale each expert's contribution, one per device. - atomic_counters (
VariadicTensors[IOSpec[_, _].MutableInput, static_specs=atomic_counters.static_specs]): Atomic counters coordinating work across thread blocks, one per device. - dev_ctxs (
DeviceContextArray): List of GPU device contexts, one per device.
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!