transformers
ec2212bd - [distributed] Rename MoE intra-expert TP styles to moe_tp_*

Commit
1 day ago
[distributed] Rename MoE intra-expert TP styles to moe_tp_* Rename registry and plan entries so TP-on-expert sharding is distinct from EP (grouped_gemm) and dense packed_colwise: moe_gate_up_colwise -> moe_tp_gate_up_colwise, moe_down_rowwise -> moe_tp_down_rowwise. Drop unused moe_tp_gate_up_colwise_alt (GPT-OSS-style layouts stay EP-only).
Author
Parents
Loading