transformers
Migrate remaining router-masked MoE models to `ep_dispatch_experts`
#49607
Open

Migrate remaining router-masked MoE models to `ep_dispatch_experts` #49607

IlyasMoutawwakil
HuggingFaceDocBuilderDev
IlyasMoutawwakil IlyasMoutawwakil changed the base branch from ep-masked-when-shared-batch to ep-passthrough-strategy 6 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the ep-passthrough-strategy branch from a7f22dbf to 1bfd6009 4 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 2b82f8bb to 5f5dc05d 4 days ago
github-actions
IlyasMoutawwakil IlyasMoutawwakil force-pushed the ep-passthrough-strategy branch from 1bfd6009 to 6ca9df45 4 days ago
IlyasMoutawwakil IlyasMoutawwakil closed this 4 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 5f5dc05d to 6ca9df45 4 days ago
IlyasMoutawwakil Move the remaining router-masked models onto `ep_dispatch_experts`
bfadfe18
IlyasMoutawwakil Drop the expert-parallel comment from the Qwen3-MoE and Mellum plans
81e3adf7
IlyasMoutawwakil IlyasMoutawwakil reopened this 4 days ago
IlyasMoutawwakil IlyasMoutawwakil changed the title Move the remaining router-masked MoE models onto `ep_dispatch_experts` Migrate remaining router-masked MoE models to `ep_dispatch_experts` 4 days ago
github-actions
vasqu
vasqu approved these changes on 2026-10-19

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone