replicate
/

moe-new-models

Model card Files Files and versions

moe-new-models / tests /kernels

Commit History

Sync with vLLM, add `apply_router_weight_on_input` arg to `fused_moe`

1aea118

danieldk HF Staff commited on Apr 7, 2025

Vendor `w8a8_block_fp8_matmul` and `per_token_group_quant_fp8`

b41d28a

danieldk HF Staff commited on Feb 4, 2025

Normalize some directory names

3d21f81

danieldk HF Staff commited on Feb 4, 2025