nvfp4-megamoe-kernel

__init__.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

cudagraph_test.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

layertest.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

test_cotiled_diag.py

…

test_cutedsl.py

…

test_d1_3_cotiled.py

D1.3: Fix while loop in cotiled diag - precompute num_tmem_alloc_cols

2026-05-24 00:07:22 +00:00

test_d1_3_layout_diag.py

…

test_d1_3_smem_diag.py

…

test_d1_3_smem_direct.py

…

test_d1_3_smem_vs_tmem.py

…

test_d1_3_unnorm_debug.py

…

test_d1_3_write_read.py

D1.3: Add SMEM-P write/read diagnostic

2026-05-24 00:13:28 +00:00

test_d1_debug.py

…

test_d1_diag2.py

…

test_d1_diag.py

…

test_d1_hd512_merge.py

…

test_d1_hd512_only.py

…

test_d1_hd512.py

…

test_d1_kv_merge_v2.py

…

test_d1_kv_merge_v3.py

…

test_d1_kv_merge.py

…

test_d1_lse_verify.py

…

test_d1_lse.py

…

test_d1_multi_kv.py

…

test_d1_qk512.py

…

test_d1_raw.py

…

test_d1_regression.py

D1.4: Fix regression test for un-normalized O output (D5a)

2026-05-24 15:13:16 +00:00

test_d1_rescale_debug.py

…

test_d1_rescale_diag.py

…

test_d1_rescale_min.py

…

test_d1_smem128.py

…

test_d1_sweep.py

…

test_d1_tmem_only.py

…

test_d1_tmem_trip.py

…

test_d2_headpacked.py

D2: comprehensive head-packed test (n_h=1, 64, 128, hd=64, 128)

2026-05-25 17:16:05 +00:00

test_d2_multicta.py

D2: add num_query_heads/batch_size params + head-packed test

2026-05-25 16:50:49 +00:00

test_d2_multihead.py

…

test_d2_perhead.py

D2: add per-head launch test

2026-05-24 22:48:22 +00:00

test_d2_regression.py

fix: use reference attn_sum for normalization (kernel LSE per-row may be wrong)

2026-05-25 17:13:34 +00:00

test_d2_scale.py

D2: add scale test (more heads, larger hd)

2026-05-24 22:49:44 +00:00

test_d3_inkernel_mask.py

fix: swa_len as Int32 scalar instead of CuTe tensor

2026-05-26 10:54:41 +00:00

test_d3_swa_mask.py

fix typo: from_dlset → from_dlpack

2026-05-25 17:28:43 +00:00

test_d4_causal_mask.py

fix: D4 test reference computation only applies causal mask when is_causal=True

2026-05-26 10:56:04 +00:00

test_d5b_perrow_lse.py

fix: k_seg is already 3D from slicing, don't add extra unsqueeze(-1)

2026-05-26 11:02:44 +00:00

test_d5c_fused.py

D5c: add apply_sink_bias flag (independent of n_comp)

2026-05-26 15:26:52 +00:00

test_d5c_multitile.py

diag: rewrite multi-tile test with explicit per-segment compile and reference

2026-05-26 15:39:39 +00:00

test_d15_in_kernel_rescale.py

D1.5: Implement in-kernel O rescale via CUTLASS correction_rescale pattern

2026-05-26 20:26:06 +00:00

test_d15_multi_kv.py

D1.5: add multi-KV-tile attention test with Python KV merge

2026-05-25 17:18:50 +00:00

test_d15_noop_rescale.py

D1.5 debug: add NOOP rescale test (acc_scale=1.0) to isolate TMEM round-trip corruption

2026-05-26 20:28:55 +00:00

test_d15_rescale_debug.py

D1.5 debug: add targeted s_k=256 rescale diagnostic test

2026-05-26 20:27:37 +00:00

test_d15_roundtrip_iso.py

D1.5: Add isolated round-trip test comparing s_k=128 vs s_k=256 with NOOP rescale

2026-05-26 20:45:58 +00:00

test_fmha_v3_stage_c.py

…

test_fmha_v3_stage_d1.py

…

test_fmha_v3_stage_d5b.py

…

test_fmha_v3.py

…

test_fp4_roundtrip.py

…

test_nvfp4_diag.py

…

test_nvfp4_gpu_quantize.py

fix test 4: use silu(gate)+swiglu interleaved (matching fused kernel output)

2026-05-25 16:24:04 +00:00

test_nvfp4_primitives.py

…

test_nvfp4_quant_kernel.py

NVFP4-1.1 Step 3: post-SWiGLU quantization test suite (all PASS)

2026-05-25 09:08:01 +00:00

test_nvfp4_quantize_kernel.py

…

test_paired_epilog.py

…

test_production.py

Stage E: production attention wrapper + Python KV merge, clean fmha_smem_acc

2026-05-27 06:34:10 +00:00

test_smem_acc.py

Add c_simple GMEM tensor (non-dynamic) for SMEM accumulator TMA store

2026-05-27 05:33:30 +00:00

test_smem_budget.py

…

test_smem_p_coord.py

…

test_smem_p_diag.py

…

test_smem_p_write.py

…

test_tmem_budget.py

…

test_tmem_roundtrip_minimal.py

D1.5: Implement in-kernel O rescale via CUTLASS correction_rescale pattern

2026-05-26 20:26:06 +00:00