veda_timing (loader, method veda) logs the time in gather, score, select, kernel and scatter
for a run; off by default and it changes no number. The run line now names the size and tile
plan of each stage, not only the last. Cheap speed-ups that keep the numbers exactly: the
list width of kept tiles comes from the layout instead of a device read, the head groups and
the per-row pattern stay on the device, and the predictor's projections are copied from pinned
memory.
Learned tile-sparse attention (Veda, arXiv 2605.30325): tiles, tile plans, predictor
scoring and selection in plain torch (tested on CPU, including a real tiny H3 with a
stand-in kernel), an INT8 Triton kernel derived from SageAttention (untested on a GPU),
installed through the same override and per-step wrapper as SLA. Runs at the predictor's
trained budget; LoRA, MXFP8, Two stages, dense steps and the attention backend compose.
Settings > Models offers the official Hugging Face download of the predictor on a click,
with its licence beside it; taggers and this share one downloader (core/fetch.py).
A few-step H3 file whose metadata keeps explicit head sizes but not the head
count was built with one head and failed on the head biases. The count is now
taken from the head bias and patch projection shapes (shapes only, any device
or dtype), with a test through the real loader at every weight dtype.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Tw9fjUsYDkyEtLCrzqv7jo
Skip fewer attention blocks on early steps and more on late ones, in a straight
line up to sla_sparsity. Off at 0, untested on H3. Credit GSRS (arXiv 2610.21129).
tools/test_env.sh and tools/cpuforce.py set up the suite on a GPU-less machine.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0176SHndmQj7T1VtWAWZiNcq
Round 1 of extensive_testing on 79ca8cf found is_h3_model() matched on
diffusion_model's bare class name ("MiniMaxH3Model"), inherited
unchanged from v4. This codebase's own core/traits.py.has_block()
documents exactly why that's unsafe for a name this generic: a
same-named class from an unrelated custom node or a future FunPack
model would be silently treated as H3 and sparsified at 0.90 with
H3-specific audio-prefix pinning -- the identical LTX-collision risk
this file's own docstring already warns about, just not guarded
against strongly enough. modules/models/minimax_h3 already qualifies
its own identical check (MODEL_CLASS = the dotted path); this port
carried over the private, weaker duplicate instead of reusing that
pattern.
Switched to core/traits.py's has_block() with the same qualified dotted
name, kept as a local string (not an import of modules/models/minimax_h3
itself, which pulls in comfy_extras at import time -- this file
deliberately stays importable without ComfyUI on the path). Added a
test exercising the exact collision the bare match would have missed:
a class named MiniMaxH3Model from an unrelated module.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Loader-level attention option: mean-pool Q and a smoothed K into
blocks, score them with one small matmul, attend only the top
1-sparsity fraction of key blocks per query block, via a vendored
Triton kernel. Runs alongside the chosen attention backend rather than
replacing it (one override slot; the backend becomes SLA's dense
fall-through for the text refiner, masked calls, and trailing dense
steps), and refuses cleanly -- naming why -- on anything that is not
MiniMax H3 or a machine without CUDA+Triton.
modules/loaders/sla_block_map.py and sla_kernel.py are near-verbatim
ports of v4's vendored files (no FunPack-specific imports, so nothing
to adapt); sla_attention.py is the loader-facing glue (H3 gate,
override composition, per-step wrapper), same content as v4's with a
"not a module" note added to match gguf_support.py's convention.
Wired into FunPackDiffusionModelLoader as six new optional widgets,
following v4's loaders.py wiring but expressed in v5's io.Schema style.
The Triton kernel itself is unexercisable on this Mac (no CUDA) --
matches how real generation is verified in this project (rented GPU,
not local). Ported the CPU-testable half of v4's test suite (the
override contract, the H3 gate, the per-step wrapper); the three
kernel-correctness tests carry the same CUDA+Triton skip guard v4 used.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
GGUF: port v4's gguf_support.py (backend selection, content-based .gguf
detection, dequantize fallback) and wire it into both the diffusion-model
and CLIP loaders.
Pipeline editor: reopening the Models-and-pipeline settings window was
silently discarding the whole edited pipeline (a loaded preset, every
loader file, every sampler setting) because GET /api/pipeline is
deliberately stateless server-side and boot.js never fed its own
already-saved slots back into a freshly remounted editor -- found live
while walking through the MiniMax-H3 setup flow. Fixed by re-validating
the saved slots through check() on reopen instead of always replaying
the raw stateless default.
Also fixes, found via the extensive_testing adversarial loop over both
of the above: a stale file-listing cache in the pipeline window (no way
to see a newly-added model file without a full page reload), and a
chain of races/leaks in its own refresh/commit error-banner handling
that surfaced across five review rounds -- a rejection from a superseded
request overwriting a fresher success, a failure banner never clearing
on success, banners bleeding across groups on navigation, and a
commit()'s refusal landing on the wrong view after the user backs out
before it answers.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
[Log] Info: FunPack Diffusion Model Loader | flux.safetensors loaded, attention sage
[Log] Alert: MiniMaxH3ImageToVideo node | no "first_image" input is connected
[Log] Warning: FunPack LoRA Loader | style.safetensors is empty and will not load
[Log] Critical: torch | cannot allocate 14 GB (3 GB free); the run will not start
The difference between the middle two is the one that matters. Alert is a
choice you may not have made on purpose. Warning is something that will NOT
do what its name says -- the level v4 needed and did not have, which is how
560 handlers came to swallow a failure and carry on, so a feature could be
switched on, do nothing, and never say so.
Info is not decoration either: "loaded on GPU using sage attention" is how
an attention setting that did nothing gets noticed months before the
pictures look wrong enough to investigate.
Every call site moved -- there is no note() left to be vague through -- and
the loaders now say what they actually loaded rather than only what they
were asked for. Records are kept in a bounded ring as well as printed, so a
long rental session cannot turn the log into a memory leak, and served at
/funpack/api/log so a person can read it somewhere other than a terminal
full of ComfyUI's own output. There is a page for it.
Also dropped the preflight this was traded for: min and max are already
declared and already enforced at both doors, and a second protection layer
around a check that exists is ceremony.
Two mistakes worth keeping: a log line reached into the model's shape to
describe it and could have failed the load it was only narrating; and both
new pages passed `children` to an element that takes `rows`, which the kit
accepted and silently dropped. It throws on an unknown element and not on an
unknown prop.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
