Aryan Kumar 21e749156c [fix]: address QAD review findings ...

- Include V2 width/height/num_frames and the train manifest in the step-0
  fingerprint so a dimension change cannot reuse a stale baseline.
- Honor CompressionSpec.first_steps in the V2 rung scope.
- Mark tagged checkpoints weights-only in metadata and teach the exporter
  to resolve and load them, so best/ can be exported.
- Add a tile-64 backward comparison against the STE reference.

2026-10-19 12:16:45 -07:00

..

__init__.py

[kernel] Reorg and fix fastvideo-kernel (#962)

2025-12-26 01:50:24 -06:00

support_flex_sta.py

[kernel] Reorg and fix fastvideo-kernel (#962)

2025-12-26 01:50:24 -06:00

test_attn_qat_infer_sparse.py

[feat]: FastH3 on single RTX GPUs (#1919)

2026-10-06 11:37:01 -07:00

test_attn_qat_infer.py

[kernel] Build + allow attn_qat_infer FP4 attention on sm_121a (DGX Spark) (#1598)

2026-10-07 04:09:39 -07:00

test_attn_qat_sm100_long_sequence.py

[perf] Default SM100 long-sequence Attn-QAT tiling (#1899)

2026-10-07 15:14:38 -07:00

test_attn_qat_train_routing.py

[perf] Default SM100 long-sequence Attn-QAT tiling (#1899)

2026-10-07 15:14:38 -07:00

test_attn_qat_train.py

[bugfix] Fix dead NVFP4 QAT plumbing: bench imports, int8 error string, smooth_q crash (#1723)

2026-09-24 17:09:33 -07:00

test_attn_qat_vsa_train.py

[fix]: address QAD review findings

2026-10-19 12:16:45 -07:00

test_block_sparse_blackwell_clc.py

[refactor] VSA: rename the sm100a block-sparse kernel to ptx_blackwell (#1956)

2026-10-17 21:10:19 -07:00

test_block_sparse_blackwell_dispatch.py

[refactor] VSA: rename the sm100a block-sparse kernel to ptx_blackwell (#1956)

2026-10-17 21:10:19 -07:00

test_block_sparse_bwd_blackwell.py

[refactor] VSA: rename the sm100a block-sparse kernel to ptx_blackwell (#1956)

2026-10-17 21:10:19 -07:00

test_build_arch.py

[kernel] Build + allow attn_qat_infer FP4 attention on sm_121a (DGX Spark) (#1598)

2026-10-07 04:09:39 -07:00

test_device_specs.py

[perf] benchmark_attn_qat_train: device peak-TFLOPS table instead of silent RTX 5090 default (#1724)

2026-09-24 17:08:03 -07:00

test_fused_compress_topk.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_nvfp4_benchmark_imports.py

[bugfix] Fix dead NVFP4 QAT plumbing: bench imports, int8 error string, smooth_q crash (#1723)

2026-09-24 17:09:33 -07:00

test_sta.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_turbodiffusion.py

[bugfix] Fix dead NVFP4 QAT plumbing: bench imports, int8 error string, smooth_q crash (#1723)

2026-09-24 17:09:33 -07:00

test_vmoba_correctness.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa128_backward.py

[perf] Add FA4 CuTe backward support for VSA-256 (#1639)

2026-08-19 11:46:49 -07:00

test_vsa256_backward.py

[perf] Add FA4 CuTe backward support for VSA-256 (#1639)

2026-08-19 11:46:49 -07:00

test_vsa256_forward_cross.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa256_forward_vbs.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa256_forward.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa256_triton.py

[perf] Add FA4 CuTe backward support for VSA-256 (#1639)

2026-08-19 11:46:49 -07:00

test_vsa_combine.py

[perf] avoid full-sequence materialization in VSA coarse/sparse combine (#1813)

2026-10-07 01:54:35 -07:00

test_vsa_forward.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa_tile_layout.py

[perf] Fuse VSA64 tile layout for SM100 inference (#1900)

2026-10-07 15:31:58 -07:00

test_vsa_triton_backward_scale.py

[bugfix] fastvideo-kernel: fix Triton block-sparse backward logit scaling (bf16 K pre-scaling) (#1730)

2026-08-21 04:50:06 -05:00

test_vsa_utils.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

test_vsa_varlen.py

[perf] Add FA4 CuTe backward support for VSA-256 (#1639)

2026-08-19 11:46:49 -07:00

test_vsa.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00

utils.py

[misc] yapf pass under CI's interpreter (3.12) + pin hook language_version (#1702)

2026-08-08 22:24:17 -07:00