sglangv0.5.2 & support Qwen3-Next-80B-A3B-Instruct

118f1fc7 · maxiao1 · 118f1fc7 · 118f1fc7 · 118f1fc7 · 118f1fc7
Commit 118f1fc7 authored Sep 13, 2025 by maxiao1
20 changed files
--- a/benchmark/kernels/deepep/deepep_utils.py
+++ b/benchmark/kernels/deepep/deepep_utils.py
--- a/benchmark/kernels/deepep/tuning_deepep.py
+++ b/benchmark/kernels/deepep/tuning_deepep.py
--- a/benchmark/kernels/deepseek/README.md
+++ b/benchmark/kernels/deepseek/README.md
--- a/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py
+++ b/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_gemm.py
--- a/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_group_gemm.py
+++ b/benchmark/kernels/deepseek/benchmark_deepgemm_fp8_group_gemm.py
--- a/benchmark/kernels/fbgemm/README.md
+++ b/benchmark/kernels/fbgemm/README.md
--- a/benchmark/kernels/fbgemm/benchmark_fbgemm_grouped_gemm.py
+++ b/benchmark/kernels/fbgemm/benchmark_fbgemm_grouped_gemm.py
--- a/benchmark/kernels/flashinfer_allreduce_fusion/README.md
+++ b/benchmark/kernels/flashinfer_allreduce_fusion/README.md
--- a/benchmark/kernels/flashinfer_allreduce_fusion/benchmark_fused_collective.py
+++ b/benchmark/kernels/flashinfer_allreduce_fusion/benchmark_fused_collective.py
--- a/benchmark/kernels/fused_moe_triton/README.md
+++ b/benchmark/kernels/fused_moe_triton/README.md
--- a/benchmark/kernels/fused_moe_triton/benchmark_sglang_fused_moe_triton.py
+++ b/benchmark/kernels/fused_moe_triton/benchmark_sglang_fused_moe_triton.py
--- a/benchmark/kernels/fused_moe_triton/benchmark_sum_scale.py
+++ b/benchmark/kernels/fused_moe_triton/benchmark_sum_scale.py
--- a/benchmark/kernels/fused_moe_triton/benchmark_torch_compile_fused_moe.py
+++ b/benchmark/kernels/fused_moe_triton/benchmark_torch_compile_fused_moe.py
--- a/benchmark/kernels/fused_moe_triton/benchmark_vllm_vs_sglang_fused_moe_triton.py
+++ b/benchmark/kernels/fused_moe_triton/benchmark_vllm_vs_sglang_fused_moe_triton.py
--- a/benchmark/kernels/fused_moe_triton/tuning_fused_moe_triton.py
+++ b/benchmark/kernels/fused_moe_triton/tuning_fused_moe_triton.py
--- a/benchmark/kernels/minmax-text-01-lightning_attention/benchmark_lightning_attention_decode.py
+++ b/benchmark/kernels/minmax-text-01-lightning_attention/benchmark_lightning_attention_decode.py
--- a/benchmark/kernels/minmax-text-01-lightning_attention/benchmark_lightning_attention_prefill.py
+++ b/benchmark/kernels/minmax-text-01-lightning_attention/benchmark_lightning_attention_prefill.py
--- a/benchmark/kernels/quantization/bench_fp4_quant.py
+++ b/benchmark/kernels/quantization/bench_fp4_quant.py
--- a/benchmark/kernels/quantization/bench_int8_quant.py
+++ b/benchmark/kernels/quantization/bench_int8_quant.py
--- a/benchmark/kernels/quantization/tuning_block_wise_kernel.py
+++ b/benchmark/kernels/quantization/tuning_block_wise_kernel.py