Ziang Li ce0541404f [FlashInfer v0.6.6][RL] Support fp8-last-n-bf16 RL for flashinfer_trtllm_routed moe backend (#20214)

2026-03-22 11:17:01 -07:00

..

compressed_tensors

Revert "[AMD] Add MoE weights and scales padding" (#21067)

2026-03-20 20:26:17 -07:00

configs

Fix result writer in tuning_block_wise_kernel.py, and add FP8 kernel config for L40 (#20368)

2026-03-20 09:28:54 +08:00

modelslim

[NPU] [Quantization] w4a4 MoE layer support (#18924)

2026-03-11 16:52:35 +03:00

quark

Revert "[AMD] Add MoE weights and scales padding" (#21067)

2026-03-20 20:26:17 -07:00

__init__.py

[Model] Support Nemotron 3 Super NVFP4 (#20407)

2026-03-14 00:56:26 -07:00

auto_round.py

[Feature][NPU]: add runtime support for AutoRound quantized models (#16699)

2026-03-07 18:03:55 +03:00

awq_triton.py

…

awq.py

Piecewise Cuda Graph set default (#16331)

2026-03-02 23:18:07 +08:00

base_config.py

[NPU] NPU quantization refactoring & more quantization formats support (#14504)

2026-01-15 04:25:15 +08:00

base_scheme.py

[2/N] Quantization Refactor: Compressed tensors MoE schemes (#17503)

2026-02-16 18:03:51 +03:00

bitsandbytes.py

Add SGLang CUDA crash API logging inspired by FlashInfer (#20910)

2026-03-22 16:39:40 +08:00

blockwise_int8.py

…

fp4_utils.py

Revert "Support CuteDSL mm_fp4 backend" (#21077)

2026-03-20 22:47:47 -07:00

fp8_kernel.py

Fix scale_step_k computation in the fp8_kernel (#20819)

2026-03-20 18:09:31 +08:00

fp8_utils.py

[AMD] Fix dpsk-v32 accuracy issue on mi355 (#20840)

2026-03-18 02:06:15 -07:00

fp8.py

Add SGLang CUDA crash API logging inspired by FlashInfer (#20910)

2026-03-22 16:39:40 +08:00

fpgemm_fp8.py

…

gguf.py

[MUSA][10/N] Add GGUF support (#18357)

2026-03-06 10:50:35 -08:00

gptq.py

[Feature][NPU]: add runtime support for GPTQ-quantized MoE models (#16364)

2026-03-04 16:02:19 +03:00

int8_kernel.py

chore: fix some typos (#18577)

2026-02-10 20:47:41 -08:00

int8_utils.py

…

kv_cache.py

Support FlashAttention3 page_size > 1 and topk > 1 case with paged attn and spec decode (#7725)

2025-11-26 11:44:41 +08:00

kvfp4_tensor.py

feat: Add FP4 (E2M1) KV Cache Support with Quantization Utilities for MLA (#10078)

2025-11-01 22:24:58 -07:00

marlin_utils_fp8.py

[PCG]add piecewise cuda graph support for marlin linear (#20119)

2026-03-11 10:57:08 +08:00

marlin_utils.py

[Kernel] Migrate GPTQ-Marlin GEMM kernel to JIT (#18067)

2026-02-06 08:31:42 +08:00

modelopt_quant.py

[Model] Support Nemotron 3 Super NVFP4 (#20407)

2026-03-14 00:56:26 -07:00

moe_wna16.py

fix_get_quant_method_in_fused_moe_condition (#18459)

2026-02-16 00:31:42 +08:00

mxfp4_tensor.py

…

mxfp4.py

Fix SM120 triton_kernels MXFP4 block_k for GPT-OSS (#20040)

2026-03-06 10:53:08 -08:00

petit_utils.py

…

petit.py

Fix circular import in quantization modules (#17372)

2026-01-21 15:47:09 +08:00

qoq.py

…

quark_int4fp8_moe.py

[AMD][Quantization] Add int4fp8_moe online quantization on ROCm (#7392)

2026-01-14 01:44:40 -08:00

rocm_mxfp4_utils.py

…

unquant.py

[FlashInfer v0.6.6][RL] Support fp8-last-n-bf16 RL for flashinfer_trtllm_routed moe backend (#20214)

2026-03-22 11:17:01 -07:00

utils.py

[RL] Support per-layer mixed FP8/BF16 serving for FP8 checkpoints (#18742)

2026-03-01 21:59:22 +08:00

w4afp8.py

Fix circular import in quantization modules (#17372)

2026-01-21 15:47:09 +08:00

w8a8_fp8.py

…

w8a8_int8.py

[CPU][INT4] Add INT4 kernels for CPU (#8226)

2026-01-29 22:30:13 -08:00