amirkl94 8b2b566ea7 Feature: Enable Flashinfer non-gated MoE bf16 (#43853) ...

Signed-off-by: Amir Klein <203507526+amirkl94@users.noreply.github.com>

2026-06-17 14:32:49 +00:00

..

configs

…

__init__.py

…

allspark_utils.py

…

flashinfer_fp4_moe.py

Deprecations for v0.23 and v0.24 (#44992)

2026-06-11 14:35:38 +00:00

flashinfer_mxint4_moe.py

…

flashinfer_utils.py

Feature: Enable Flashinfer non-gated MoE bf16 (#43853)

2026-06-17 14:32:49 +00:00

fp8_utils.py

[Model] Add MiniMax M3 support (#45381)

2026-06-16 01:01:25 +08:00

gptq_utils.py

Revert "[MoE Refactor] Migrate MoeWNA16Method quantization to MK orac… (#44033)

2026-05-29 16:45:29 -07:00

humming_utils.py

[Bugfix] Lazily import the humming quantization backend (#44921)

2026-06-10 06:06:17 -07:00

int8_utils.py

Enable perf_token_group_quant/_C_stable_libtorch for ROCm (#42758)

2026-06-02 23:23:28 -07:00

layer_utils.py

…

machete_utils.py

…

marlin_utils_fp4.py

[Kernel] Consolidate Marlin thread-tile padding across all dense Marlin paths (#45295)

2026-06-12 13:46:21 -07:00

marlin_utils_fp8.py

[Kernel] Consolidate Marlin thread-tile padding across all dense Marlin paths (#45295)

2026-06-12 13:46:21 -07:00

marlin_utils_test.py

…

marlin_utils.py

[Kernel] Consolidate Marlin thread-tile padding across all dense Marlin paths (#45295)

2026-06-12 13:46:21 -07:00

mxfp4_utils.py

[ROCm][gpt-oss] Hybrid CDNA4 swizzle gate for A8W4 MoE (#44804)

2026-06-09 23:59:44 -07:00

mxfp6_utils.py

…

mxfp8_utils.py

[Model] Add MiniMax M3 support (#45381)

2026-06-16 01:01:25 +08:00

nvfp4_emulation_utils.py

[NVFP4][Hopper/AMD Instinct] Add Triton kernels for NVFP4 dequantization and QDQ emulation (#40033)

2026-04-30 17:35:48 -04:00

nvfp4_utils.py

…

ocp_mx_utils.py

…

quant_utils.py

[Refactor] Remove Fp8OnlineLinearMethod as scheduled (#45463)

2026-06-16 04:35:58 -07:00

w8a8_utils.py

…