Trevor Morris e9f8e42318 Support FP4 gemm (1/2) (#3899)

2025-03-24 19:50:23 -07:00

..

speculative

Add greedy verification kernel (#4383)

2025-03-16 00:58:26 -07:00

test_activation.py

sync flashinfer and update sgl-kernel tests (#3081)

2025-01-23 21:13:55 +08:00

test_awq_dequant.py

[1/3] fix dsv3 awq issue (#4556)

2025-03-22 01:07:17 -07:00

test_bmm_fp8.py

feat: integrate bmm_fp8 kernel into sgl-kernel (#3056)

2025-01-23 00:39:38 +08:00

test_cublas_grouped_gemm.py

[Feature] Apply Cublas Grouped Gemm kernel (#3629)

2025-02-18 15:18:31 +08:00

test_deep_gemm.py

DeepGemm integrate to sgl-kernel (#4165)

2025-03-10 00:35:07 -07:00

test_fp4_gemm.py

Support FP4 gemm (1/2) (#3899)

2025-03-24 19:50:23 -07:00

test_fp4_quantize.py

Support FP4 gemm (1/2) (#3899)

2025-03-24 19:50:23 -07:00

test_fp8_blockwise_gemm.py

support blockwise fp8 matmul kernel (#3267)

2025-02-13 01:49:33 +08:00

test_fp8_gemm.py

support w8a8 fp8 kernel with CUTLASS (#3047)

2025-01-26 15:46:51 +08:00

test_int8_gemm.py

Support serving DeepSeek-R1-Channel-INT8 with 32 L40S. (#4418)

2025-03-17 00:03:43 -07:00

test_lightning_attention_decode.py

[Fix] use torch.cat instead of torch.concat to prevent entering the Autograd backends. (#4466)

2025-03-16 00:02:47 -07:00

test_moe_align.py

Remove the vllm dependency from the moe_align function (#4164)

2025-03-07 22:42:16 -08:00

test_moe_topk_softmax.py

Add moe topk softmax templated from vllm (#4302)

2025-03-14 12:03:33 -07:00

test_norm.py

cleanup sgl-kernel kernels (#3175)

2025-01-27 19:11:01 +08:00

test_per_tensor_quant_fp8.py

[quant kernel] sgl-kernel support per_tensor_quant fp8 (#3786)

2025-03-06 18:05:43 -08:00

test_per_token_group_quant_8bit.py

[Quant Kernel] refactored per token group quant fp8 to support int8 up-to 2x faster (#4396)

2025-03-23 23:44:17 -07:00

test_per_token_quant_fp8.py

fix accuracy issue (#4376)

2025-03-13 02:06:22 -07:00

test_rotary_embedding.py

Simplify tests & Fix trtllm custom allreduce registration (#4252)

2025-03-10 01:24:22 -07:00

test_sampling.py

feat: integrate sampling kernels into sgl-kernel (#3086)

2025-01-24 01:54:47 +08:00

test_trt_allreduce.py

Rename files in sgl kernel to avoid nested folder structure (#4213)

2025-03-08 22:54:51 -08:00