Xiaoyu Zhang 1360848ee1 Optimize large GroupNorm SiLU apply (#23938)

2026-05-02 20:54:46 +08:00

..

diffusion

Optimize large GroupNorm SiLU apply (#23938)

2026-05-02 20:54:46 +08:00

bench_activation.py

[MoE] Deprecate act_and_mul_triton; fold filter_expert into JIT silu/gelu_and_mul (#23707)

2026-04-26 01:41:35 -07:00

bench_awq_dequantize.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_cast.py

Remove obsolete sgl-kernel legacy paths (#21528)

2026-04-01 09:00:20 +08:00

bench_clamp_position.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_concat_mla.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_custom_all_reduce.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_fused_qknorm_rope.py

[jit_kernel] Optimize fused_qknorm_rope: deduplicate sincosf for interleave RoPE (#21654)

2026-04-01 09:04:13 +08:00

bench_hadamard.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_hicache.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_hisparse.py

[HiSparse]: Add benchmark for hisparse kernel (#22187)

2026-04-13 12:49:18 +08:00

bench_mxfp8_moe.py

[JIT Kernel][1/2]Migrate MXFP8 Group GEMM & Quant into JIT (#23833)

2026-04-29 22:50:09 +08:00

bench_norm.py

[Feature] JIT rmsnorm update (with claude) (#21834)

2026-04-01 23:40:00 +08:00

bench_nvfp4_blockwise_moe.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_nvfp4_quant.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_nvfp4_scaled_mm.py

CUTLASS NVFP4 GEMM improvement of SM120 (#21314)

2026-04-01 09:04:34 +08:00

bench_per_tensor_quant_fp8.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_per_token_group_quant_8bit.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_qknorm_across_heads.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_qknorm.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_renorm.py

Remove obsolete sgl-kernel legacy paths (#21528)

2026-04-01 09:00:20 +08:00

bench_resolve_future_token_ids.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_rope.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_store_cache.py

Refactor JIT kernel CI to use run_suite.py registration system (#21239)

2026-03-23 21:17:27 -07:00

bench_tp_qknorm.py

[Feature][JIT Kernel] Fused TP QK norm For Minimax (#20673)

2026-04-13 20:29:47 +08:00

utils.py

[Feature] JIT rmsnorm update (with claude) (#21834)

2026-04-01 23:40:00 +08:00