kkandwunhuang a6001478f4 [AMD] Perf Kimi-K3 MoE optimization (#33838) ...

Co-authored-by: wunhuang <wunhuang@amd.com>

2026-09-03 02:28:28 -07:00

..

test_compressed_tensors_kv_cache.py

[Quant] Load compressed-tensors kv_cache_scheme scales (#35455)

2026-08-20 19:17:59 +08:00

test_compressed_tensors_lm_head.py

[Quant] Load compressed-tensors quantized lm_head instead of value-casting it (#35228)

2026-08-19 15:37:45 -07:00

test_compressed_tensors_mixed_precision.py

[Bugfix] compressed-tensors: mixed-precision checkpoints silently load unquantized (#32736)

2026-07-29 09:55:33 -07:00

test_compressed_tensors_wna16_moe_no_linear.py

[Config] Round 5.1: the published-side readers ask the bags, and a platform fact gets one address (#37086)

2026-08-30 02:18:33 -07:00

test_deepgemm_ue8m0_requant.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

test_flashinfer_pr4266_bf16_gemm.py

qwen 3.8 rebase (#35758)

2026-08-28 20:41:34 -07:00

test_flashinfer_trtllm_fp8_fallback.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

test_fp4_kv_cache_quant_method.py

[CPU] Support FP8 KV cache (#32733)

2026-09-02 10:20:54 +08:00

test_fp8_blockwise_linear_backends.py

add flashinfer cute-dsl backend for mxfp8 gemm (#34042)

2026-08-13 08:50:01 +08:00

test_fp8_kernel_hip_max.py

[AMD][Quantization][Bugfix] Fix bug related to fp8 max on gfx95x for per-token-group quant (ROCm) (#30900)

2026-08-15 19:54:16 -07:00

test_fp8_moe_runner_ownership.py

fix(moe): guard FP8 delegate activation params (#36275)

2026-08-26 12:12:21 +00:00

test_fp8_moe_weight_gating.py

Fix MXFP8 MoE weight sizing for non-gated models (#36097)

2026-08-25 22:09:09 +08:00

test_fp8_utils_mxfp4.py

[NVIDIA] Fix SM107 MXFP8 activation prep (#35405)

2026-08-23 18:17:03 +08:00

test_gptq_scheme_attach.py

[Quantization] Fix GPTQ scheme attachment broken by LinearBase.scheme default (#34962)

2026-08-16 00:48:02 -07:00

test_humming_w4afp8_schemas.py

feat(humming): support native W4AFP8 checkpoint schemas (#32033)

2026-08-24 18:59:27 +08:00

test_int8_linear_methods.py

config: a parallel size has one spelling; a patched scope declares its own (#36621)

2026-08-27 12:56:42 -07:00

test_marlin_utils_fp8.py

[Fix] Correct dense FP8 Marlin bias ordering (#35020)

2026-08-17 03:43:36 +00:00

test_modelopt_nvfp4_moe_scales.py

[Fix] Support Kimi-K3 ModelOpt mixed NVFP4/FP8 checkpoint (#35077)

2026-08-19 08:13:45 -07:00

test_modelopt_nvfp4.py

[ModelOpt FP4] Support online MoE weight quantization (#33115)

2026-08-06 11:01:55 -07:00

test_mxfp4_flashinfer_activation_prep.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

test_mxfp4_situ_output.py

[AMD] Perf Kimi-K3 MoE optimization (#33838)

2026-09-03 02:28:28 -07:00

test_mxfp4_situ_weight_layout.py

[AMD] Perf Kimi-K3 MoE optimization (#33838)

2026-09-03 02:28:28 -07:00

test_mxfp4_sm90_cutlass.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

test_mxfp4_sm120_cutlass.py

[MoE] Add FlashInfer SM90 MXFP4 W4A8 CUTLASS MoE (#34967)

2026-08-31 20:04:41 -07:00

test_nvfp4_linear_backends.py

[Test] Route GEMM backend UTs through real layer modules and weight loaders (#33615)

2026-08-04 20:53:26 -07:00

test_nvfp4_moe_backends.py

[CI][RFC] Replace black-jupyter with ruff-format (#37210)

2026-09-02 19:46:08 -07:00

test_quark_config.py

[AMD][Quantization] Online MXFP4 quantization 4/N - NVFP4 to MXFP4 Online Requantization on AMD GPUs (#29328)

2026-08-14 21:59:39 -07:00

test_quark_utils.py

[misc] Remove unit test cases that fail the admission criteria (round 3) (#30713)

2026-07-09 19:42:51 -07:00