Brayden Zhong 6a9b09847c CUTLASS NVFP4 GEMM improvement of SM120 (#21314)

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

..

bench_activation.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_amd_deterministic_allreduce.py

[amd] Add deterministic all-reduce kernel for AMD (ROCm) (#15340)

2025-12-18 23:36:03 -08:00

bench_awq_dequant.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_cutlass_mla.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_dsv3_fused_a_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_dsv3_router_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_es_fp8_blockwise_grouped_gemm.py

[sgl-kernel][6/7]Support Expert Specialization Grouped GEMM (#15471)

2026-03-19 15:39:52 +08:00

bench_fp4_gemm.py

CUTLASS NVFP4 GEMM improvement of SM120 (#21314)

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

bench_fp8_blockwise_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_fp8_blockwise_group_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_fp8_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_int8_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_kimi_k2_moe_fused_gate.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_moe_align_block_size.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_moe_ep_post_reorder.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_moe_fused_gate.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_moe_topk_sigmoid.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_moe_topk_softmax.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_mrope.py

Support mrope triton kernel and add unit test (#11722)

2025-10-20 11:51:07 +08:00

bench_per_tensor_quant_fp8.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_per_token_group_quant_8bit.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_per_token_quant_fp8.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_qserve_w4a8_gemm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_rmsnorm.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_rotary_embedding.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_sum_scale.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00

bench_top_k_top_p_sampling.py

refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009)

2026-03-20 05:55:36 -07:00