William Lin
552d7dfadc
[perf] PAN2 sampling: Wan2.2 VAE decode kernel, split decode, fused DiT kernels
...
- fastvideo-kernel/wan22_vae_decode: the Wan2.2 (TI2V-5B) VAE decoder on sm_100a / sm_103a at 704x1248, built on
first use against fastvideo-kernel's CUTLASS submodule. It frees intermediates early and carries 4 latent frames
per pass: a 5 s video peaks at 20.3 GiB, down from 94.6 GiB in the kernel's first version, with the same output.
- WanVAEConfig.decode_kernel routes AutoencoderKLWan.decode through it where it applies (bf16 with fp32
accumulation, about 7x faster than the fp32 decode on a B300, PSNR 61-63 dB, SSIM 0.998 against it);
decode_kernel_parallel splits the decode over the sequence-parallel group (2, 4 or 8 GPUs) and broadcasts the
video, with the ranks agreeing on whether every one of them could build the kernel.
- PAN2's DiT runs fused Triton kernels for Q/K RMSNorm + RoPE and LayerNorm + modulation (pan2_fusions,
set_fused_kernels).
- PAN2PipelineConfig turns all three on (vae_decode_kernel, vae_decode_kernel_parallel, dit_fused_kernels).
- VideoGenerator copies a decoded RGB video's frames to the host once instead of once per frame.
- The PAN2 presets sample at guidance 2.5, the reference launcher's default since 2026-10-23.
2026-11-02 21:42:30 +00:00
William Lin
3fb291b496
[new-model] Re-add PAN2 text-to-video and image-to-video (reverts #2025 )
...
This reverts commit 6743f4b4b#1961 .
2026-11-02 21:42:30 +00:00
4f80c5a3da
[feat] Open-H3-Regenerate-2K: regenerate FastH3 QAD, Trim and V2 drafts at 2K (text to 2K 3.1-3.4x faster) ( #2037 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: Aryan <aryan@aryan5v.dev >
2026-11-02 03:44:28 -08:00
Aryan Kumar and Aryan Kumar
b0047cfedc
[feat] Open-H3-Regenerate-2K: NVFP4 H3 video decoder (decode 23 s to 5 s at 2K) ( #2036 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-11-01 22:51:41 -08:00
Aryan Kumar and Aryan Kumar
ac81198d33
[new-model] Open-H3-Regenerate-2K: 768P H3 draft to 2560x1440, served on the OpenAI video routes ( #2035 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-11-01 22:45:09 -08:00
Aryan Kumar and Aryan Kumar
cb02062930
[new-model] Add Pika H3 Flex inference support ( #2038 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-11-01 22:33:10 -08:00
Aryan Kumar and Aryan Kumar
d954fad9ec
[perf] MiniMax H3: faster FP4 attention padding, uint8 VAE pixel path and cuDNN FP4 GEMMs ( #2046 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-11-01 22:23:10 -08:00
Aryan Kumar and Aryan Kumar
4d0a10d989
[docs] Fix the docs build, cookbook memory status, and broken commands ( #2023 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-10-30 17:08:41 -07:00
Aryan Kumar
6743f4b4b7
Revert "[new-model] Add PAN2 text-to-video and image-to-video" ( #2025 )
2026-10-27 18:03:59 -07:00
William Lin and aryan5v
a0e6cf1f50
[new-model] Add PAN2 text-to-video and image-to-video ( #1961 )
...
Co-authored-by: aryan5v <email.aryan1005@gmail.com >
2026-10-25 20:20:00 -07:00
Loay Rashid and Aryan Kumar
0d05b09c57
[new-model] Add initial Qwen-Image-2.1 inference ( #1918 )
...
Co-authored-by: Aryan Kumar <156166681+aryan5v@users.noreply.github.com >
2026-10-25 17:28:49 +00:00
posiposi443-ops and Aryan Kumar
8db2d71ef2
[perf] Cache MLX H3 text refinement and speed up attention ( #2002 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-10-23 12:40:47 -07:00
Aryan Kumar and Aryan Kumar
fa2d1324c6
[feat]: FastH3 Omni (Ref2VA) on Apple Silicon MLX ( #2000 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-10-21 12:01:54 -07:00
Aryan Kumar and Aryan Kumar
089a5a55a0
[feat] Cookbook serving, hardware catalog, and landing page ( #1999 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-10-20 09:10:29 -07:00
376e445f0e
[feat]: FastH3 OmniRef inference on consumer devices (RTX 5090 / PRO 6000 / DGX Spark / 4090) ( #1980 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com >
2026-10-19 14:57:36 -07:00
2099dd5533
[feat]: Community MiniMax-H3 LoRA loading with quantized side branch ( #1973 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
2026-10-19 14:29:23 -07:00
lpc0220 and root
d4ed61cc7c
[refactor] VSA: rename the sm100a block-sparse kernel to ptx_blackwell ( #1956 )
...
Co-authored-by: root <root@ptyche0124.ptyche.clusters.nvidia.com >
2026-10-17 21:10:19 -07:00
Aryan Kumar and Aryan Kumar
6212e7262e
[docs] Migrate documentation and cookbook to Blume ( #1969 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-10-16 16:09:04 -07:00
William Lin
9f59519959
[misc] Remove DreamX World, HunyuanGameCraft, GLM-Image, LongCat, TurboDiffusion, and Z-Image ( #1926 )
2026-10-14 23:16:32 -07:00
8530fdc550
[feat] Add external-launcher executor for multi-node inference ( #1746 )
...
Co-authored-by: Will Lin <willlin@nvidia.com >
Co-authored-by: William Lin <8941107+SolitaryThinker@users.noreply.github.com >
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 04:46:26 -07:00
Raghav K and Claude Opus 5.5
315b25d26d
[new-model] Add Cosmos Predict2.5 distilled T2W and DFD V2W inference ( #1767 )
...
Adds the Cosmos Predict 2.5 distilled text-to-world sampler (sCM, per-frame timesteps) and the DFD video-to-world pipeline, with their schedulers, registry entries, presets and examples.
Maintainer review fixes (Swarm orders #718 and #748 ): the DiT receives one shared fps value (its RoPE table is batch-independent, so per-sample vectors crashed batch > 1; differing values now raise), non-finite rollouts raise FloatingPointError instead of being zeroed by nan_to_num, the DFD detector matches "dfd" only in the last path component, the distilled example defaults to 24 fps, and the distilled rollout keeps its state in FP64 through the scheduler step as the official driver does.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 04:42:44 -07:00
af78c759f1
[new-model] Add Wan2.2-Animate-14B character animation/replacement inference ( #1765 )
...
Co-authored-by: SuhaanCommits <suhaan@cedar.build >
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
2026-10-07 04:19:37 -07:00
f82c190cbb
[kernel] Build + allow attn_qat_infer FP4 attention on sm_121a (DGX Spark) ( #1598 )
...
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 04:09:39 -07:00
2d6e7d9b6b
[feat]: MiniMax-H3 encoder split: dedicated text-encoder node group for 720p on DGX Spark ( #1877 )
...
Co-authored-by: magicbear <magicbear@users.noreply.github.com >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 03:46:33 -07:00
William Lin and Claude Opus 5.5
6d6195a9f1
[feat]: fp8 PV mode for the FA4-FP4 attention path ( #1654 )
...
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 03:36:54 -07:00
2452022f15
[new-model] Add LTX-2.5 inference support ( #1704 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: CodeRabbit <noreply@coderabbit.ai >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 03:34:18 -07:00
4b3f99b224
[new-model] Added LingBot-World-Fast image-to-video support ( #1665 )
...
Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
2026-10-07 03:32:38 -07:00
6e4813f8b4
[bugfix] use in-process executor for num_gpus=1 ( #1755 ) ( #1879 )
...
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com >
2026-10-07 02:18:23 -07:00
ef62ac48d1
[new-model] Add Wan2.2-S2V-14B speech-to-video (model port + checkpoint converter) ( #1683 )
...
Co-authored-by: SuhaanCommits <suhaan@cedar.build >
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
2026-10-07 01:36:01 -07:00
William Lin
33d81730e9
[feat]: Kandinsky 6 Lite: register the 3.2B checkpoints and add docs + cookbook recipes ( #1927 )
2026-10-06 17:09:53 -07:00
8b51380466
[feat]: FastH3 on single RTX GPUs ( #1919 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: Aryan Kumar <aryank@Aryans-Mac-Studio.local >
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com >
2026-10-06 11:37:01 -07:00
f06d824054
[feat]: add Kandinsky-6.0 TI2VA and VSR pipelines ( #1924 )
...
Co-authored-by: leffff <levnovitskiy@gmail.com >
Co-authored-by: Kirill <kykozlov@gmail.com >
Co-authored-by: shaoxiongduan <shaoxiongduan@gmail.com >
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com >
2026-10-06 10:50:17 -07:00
8444c0897a
[feat] FastH3 Ref2VA PDD inference with reference-video VSA sparsity ( #1907 )
...
Co-authored-by: Davids048 <jundasu@ucsd.edu >
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com >
2026-10-02 11:28:07 -07:00
Junda Su
9edc8adf5f
[misc] Move test-suite environment access into the registry and allowlists ( #1898 )
2026-09-30 00:15:39 -07:00
Aryan Kumar and Aryan Kumar
7f03e03dc6
[docs]: FastH3 V2 cookbook serving and MLX install guide ( #1884 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-09-29 13:03:18 -07:00
Aryan Kumar
361f919c88
[feat]: add a narrow FastH3 8-step MLX ladder
...
Keep the four-step uniform AdaLN cache, and use explicit DMD rungs only when the snapshot declares its own scheduler shifts.
2026-09-23 11:06:55 -07:00
Keith and SolitaryThinker
c4824c7764
[bugfix] FP8: gate the ROCm _scaled_mm path on CDNA4 (gfx950) instead of the capability tuple ( #1859 )
...
Co-authored-by: SolitaryThinker <wlsaidhi@gmail.com >
2026-09-16 19:58:55 -07:00
William Lin
0100218594
[feat] Support the FastH3 8-Step V2 checkpoint: checkpoint-defined shifts, explicit DMD schedule, new example ( #1852 )
2026-09-15 14:13:45 -07:00
William Lin
556ac7088e
[refactor] Simplify Wan sampling and tests ( #1825 )
2026-09-07 15:57:19 -07:00
William Lin
e19913f6e9
[refactor] Group Wan transformer and config ( #1823 )
2026-09-05 17:05:32 -07:00
aaef49bfc6
[feat]: run FastH3 across two DGX Sparks with Ray sequence parallel ( #1803 )
...
Co-authored-by: Kyle <shh075@ucsd.edu >
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com >
Co-authored-by: Satyam Srivastava <srivastavasatyam53@gmail.com >
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-09-01 01:37:23 -07:00
Aryan Kumar and Aryan Kumar
a28f2bab4b
[feat] Add an optional MLX TAEH3 preview decoder ( #1794 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-08-31 05:08:06 -07:00
Aryan Kumar and Aryan Kumar
f82d8be4bf
[perf] Sequential MiniMax H3 start with GPU-direct DiT load ( #1793 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-08-31 04:20:21 -07:00
Aryan Kumar and Aryan Kumar
620bc36dc4
[docs]: Cookbook catalog improvements ( #1790 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-08-30 17:50:57 -07:00
Suhaan Khurana
29ff16ec96
[feat] Add MiniMax H3 MLX spatial fast mode ( #1789 )
2026-08-30 17:47:02 -07:00
a4d9a75e2c
[perf] Add MiniMax H3 MLX VSA and SIMD attention ( #1776 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
2026-08-30 13:44:54 -07:00
ac48bb3cd1
[feat] Add MiniMax H3 MLX T2VA inference ( #1770 )
...
Co-authored-by: Aryan Kumar <aryank@Aryans-Mac-Studio.local >
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: CodeRabbit <noreply@coderabbit.ai >
2026-08-28 12:54:33 -07:00
KyleNeverGivesUp
e9bbaca07d
[perf] Disable every offload path on unified memory, unblocking MiniMax H3 generation on one GB10 ( #1715 )
2026-08-26 15:32:56 -07:00
Aryan Kumar and Aryan Kumar
6388db815b
[bugfix] FastMetal-QAD MLX support: refuse CUDA QAD trees, use packed mlx_dit config, stream loads ( #1736 ) ( #1758 )
...
Co-authored-by: Aryan Kumar <aryan5v@users.noreply.github.com >
2026-08-25 14:51:40 -07:00
William Lin
f9e3680f11
[perf] Align FastH3 optimized inference profile ( #1748 )
2026-08-23 02:01:03 -07:00
William Lin