Ming Yang 504d914314 [Perf] Add Triton config for DeepSeek V3 FP8 EP32 H200 (#23504) ...

Signed-off-by: Ming Yang <minos.future@gmail.com>

2025-08-24 18:06:35 -07:00

..

configs

[Perf] Add Triton config for DeepSeek V3 FP8 EP32 H200 (#23504)

2025-08-24 18:06:35 -07:00

__init__.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

allspark_utils.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

bitblas_utils.py

[Bugfix] Add proper comparison for package versions (#22314)

2025-08-06 20:31:03 -07:00

flashinfer_fp4_moe.py

[Kernels] Clean up FusedMoeMethodBase and modular kernel setup. Remove extra arguments from modular kernel methods. (#22035)

2025-08-15 14:46:00 -04:00

flashinfer_utils.py

[NVIDIA] Add SM100 Flashinfer Cutlass MoE fp8 backend (#22357)

2025-08-19 18:01:53 -04:00

fp8_utils.py

[Feature] Enable DeepGEMM Linear on B200; 1.5% E2E throughput improvement (#23351)

2025-08-21 21:01:08 -07:00

gptq_utils.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

int8_utils.py

[Perf] Cuda Kernel for Int8 Per Token Group Quant (#21476)

2025-07-25 17:07:07 -07:00

layer_utils.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

machete_utils.py

Enable group size 64 for Machete (#20290)

2025-07-01 18:05:44 -07:00

marlin_utils_fp4.py

[Kernel] [Quantization] Add MXFP4 and bias support for marlin kernel (#22428)

2025-08-14 11:23:22 -07:00

marlin_utils_fp8.py

[Kernel] [Quantization] Add MXFP4 and bias support for marlin kernel (#22428)

2025-08-14 11:23:22 -07:00

marlin_utils_test_24.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

marlin_utils_test.py

[Misc] Add SPDX-FileCopyrightText (#19100)

2025-06-03 11:20:17 -07:00

marlin_utils.py

[Kernel] [Quantization] Add MXFP4 and bias support for marlin kernel (#22428)

2025-08-14 11:23:22 -07:00

mxfp4_utils.py

[Kernel] Add cuda kernel for gpt_oss activation (#22951)

2025-08-17 05:03:24 +00:00

nvfp4_emulation_utils.py

[Refactor] Refactor MOE NVFP4 Code Base: ModelOpt + Compressed Tensor (#21631)

2025-07-27 05:25:21 -07:00

nvfp4_moe_support.py

Support Tensorrt-LLM MoE fp4 for low-latency (#21331)

2025-08-07 19:18:22 -07:00

petit_utils.py

Quantization: support FP4 quantized models on AMD CDNA2/CDNA3 GPUs (#22527)

2025-08-22 20:53:21 -06:00

quant_utils.py

[NVIDIA][torch.compile] Support Flashinfer TRTLLM FP8-q/kv NVFP4-out Attention Kernel (#22703)

2025-08-22 22:09:05 +00:00

w8a8_utils.py

[BugFix][AMD][Quantization] Fix torch.compile issue where wvSplitKQ not being called when it should when using quantized FP8 model (#22281)

2025-08-22 21:47:57 +00:00