Cheng Yanfei fbcccaa750 [Intel HPU] enable MoE EP for hpu (#5855) ...

* enable HPU MoE EP

* MoE intermediate_scale stack

* enable loader_v1 esp for tensor_wise_fp8 TP or EP

* modify activation_scale name

2026-01-15 13:08:00 +08:00

..

__init__.py

support w4afp8 EP inference (#3044)

2025-08-25 11:27:45 +08:00

ep.py

[Feature] Support redundant expert for eplb (#5918)

2026-01-09 17:13:24 +08:00

fused_moe_backend_base.py

[Feature] Support redundant expert for eplb (#5918)

2026-01-09 17:13:24 +08:00

fused_moe_cutlass_backend.py

[BugFix] fix w4afp8 tp=8 (#5868)

2026-01-05 18:59:02 +08:00

fused_moe_deepgemm_backend.py

add m_grouped_gemm_fp8_fp8_bf16_nt_contiguous_custom_python_op (#5847)

2026-01-07 16:17:55 +08:00

fused_moe_marlin_backend.py

[New][RL] Support Rollout Routing Replay (#5405)

2025-12-05 22:06:26 +08:00

fused_moe_triton_backend.py

[GraphOptimization] Wrap deep gemm and triton as python op (#5673)

2025-12-24 15:23:46 +08:00

fused_moe_wint2_backend.py

[New][RL] Support Rollout Routing Replay (#5405)

2025-12-05 22:06:26 +08:00

moe.py

[Intel HPU] enable MoE EP for hpu (#5855)

2026-01-15 13:08:00 +08:00

routing_indices_cache.py

[RL][CI] Support Async R3 And Add Accuracy Test (#5937)

2026-01-14 04:25:06 -08:00

triton_moe_kernels.py

[OPs] MoE support wfp8afp8(channelwise) and improve per_token_quant_fp8 (#4238)

2025-09-24 16:39:51 +08:00