fzyzcjy e85cb1ce9d Fix quant kernel test errors and benchmark wrong output speeds (#7604)

2025-08-21 03:48:41 -07:00

..

spatial

…

speculative

…

test_activation.py

…

test_apply_token_bitmask_inplace.py

fix sgl-kernel unit tests (#5666)

2025-04-23 01:18:30 -07:00

test_awq_dequant.py

…

test_bmm_fp8.py

…

test_custom_allreduce.py

…

test_cutlass_mla.py

…

test_cutlass_w4a8_moe_mm.py

…

test_dsv3_fused_a_gemm.py

…

test_dsv3_router_gemm.py

[Kimi K2] dsv3_router_gemm supports NUM_EXPERTS == 384 (#8013)

2025-08-01 22:01:24 +08:00

test_ep_moe_post_reorder_kernel.py

…

test_ep_moe_pre_reorder_kernel.py

…

test_ep_moe_silu_and_mul_kernel.py

…

test_flash_attention.py

Reduce overhead for fa by not calling heavy CUDA property check (#7375)

2025-08-20 16:26:28 +08:00

test_fp4_gemm.py

…

test_fp4_quantize.py

…

test_fp8_blockwise_gemm.py

…

test_fp8_blockwise_moe.py

…

test_fp8_gemm.py

…

test_gptq_kernel.py

[2/n]decouple quantization implementation from vLLM dependency (#8112)

2025-08-14 03:19:03 -07:00

test_int8_gemm.py

…

test_kvcacheio.py

…

test_lightning_attention_decode.py

[Misc] Clean sgl-kernel test (#5216)

2025-04-10 11:28:41 -07:00

test_marlin_gemm.py

[2/n]decouple quantization implementation from vLLM dependency (#8112)

2025-08-14 03:19:03 -07:00

test_marlin_repack.py

[2/n]decouple quantization implementation from vLLM dependency (#8112)

2025-08-14 03:19:03 -07:00

test_merge_state_v2.py

…

test_merge_state.py

…

test_moe_align.py

…

test_moe_fused_gate.py

[2/2] Fuse routed scaling factor into select_experts (#8690)

2025-08-20 15:10:16 -07:00

test_moe_topk_softmax.py

…

test_mscclpp.py

…

test_norm.py

…

test_per_tensor_quant_fp8.py

…

test_per_token_group_quant_8bit.py

Fix quant kernel test errors and benchmark wrong output speeds (#7604)

2025-08-21 03:48:41 -07:00

test_per_token_quant_fp8.py

fix per token cuda kernel hidden dim cannot divide by 16 (#8543)

2025-08-01 09:27:18 -07:00

test_qserve_w4a8_per_chn_gemm.py

…

test_qserve_w4a8_per_group_gemm.py

…

test_rotary_embedding.py

Fuse writing KV buffer into rope kernel (part 1: sgl-kernel) (#9077)

2025-08-12 01:46:40 -07:00

test_sampling.py

[sgl-kernel] Support FlashInfer top_k_top_p_sampling_from_logits (#9060)

2025-08-14 10:56:36 -07:00

test_sparse_flash_attn.py

…