sglang_v0.5.2/flashinfer_0.3.1/csrc
hailin 06e45b5ff9 local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
..
fused_moe/cutlass_backend local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
nv_internal local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
xqa local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
activation.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
aot_extension_utils.h local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_attention.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_attention_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_attention_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_attention_paged_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_mla_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_mla_cute_sm80.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_mla_plan.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_mla_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_decode_mla_run.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_plan.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_run.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_sm90_plan.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_sm90_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_mla_sm90_run.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_fp8_paged_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_fp8_ragged_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_fp8_sm90.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_paged_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_paged_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_ragged_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_ragged_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_sm90.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_sm90_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_sm90_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
batch_prefill_sm90_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
blackwell_fmha_plan.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
bmm_fp8.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
cascade.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
cudnn_sdpa_kernel_launcher.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
cudnn_sdpa_utils.h local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
cutlass_mla.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_cascade_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_gemm_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_gemm_sm90_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_mla_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_norm_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_page_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_quantization_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_rope_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_sampling_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
flashinfer_xqa_ops.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fmha_cutlass_sm100.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fmha_cutlass_sm100_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp4_gemm_cutlass.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp4_gemm_cutlass.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp4_gemm_cutlass_sm120.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp4_gemm_cutlass_sm120.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp8_gemm_cutlass.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
fp8_gemm_cutlass.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_groupwise_sm100.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_groupwise_sm100_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_groupwise_sm120.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_groupwise_sm120_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_sm100_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
gemm_sm120_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_fp8_groupwise_sm100.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_fp8_groupwise_sm100_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_fp8_groupwise_sm120.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_fp8_groupwise_sm120_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_mxfp4_groupwise_sm100.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_mxfp4_groupwise_sm100_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_sm90.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_sm100_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
group_gemm_sm120_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
logging.cc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
norm.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
nvshmem_binding.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
page.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pod.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pod_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pod_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pod_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pod_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pytorch_conversion_utils.h local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
pytorch_extension_utils.h local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
quantization.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
renorm.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
rope.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
runtime_utils.h local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
sampling.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_decode.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_decode_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_decode_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_decode_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_decode_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_fp8_sm90.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_fp8_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_sm90.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_sm90_config.inc local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_sm90_customize_config.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_sm90_jit_pybind.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
single_prefill_sm90_kernel_inst.jinja local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_allreduce.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_allreduce_fusion.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_alltoall.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_alltoall_prepare.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_batched_gemm_runner.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fmha_kernel_launcher.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_dev_kernel.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_kernel_launcher.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_routing_deepseek.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_routing_llama4.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_routing_renormalize.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_fused_moe_runner.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_gemm_runner.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_mnnvl_allreduce.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
trtllm_moe_allreduce_fusion.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00
vllm_custom_all_reduce.cu local source code for flashinfer_0.3.1 && torch vision_0.22.1 2025-09-20 12:33:58 +08:00