Spaces:
Running on Zero
Running on Zero
Commit History
3 days ago
-
docs: every sentence stays true after the deploy: README and the release notes say this is deploy 2's code and point to issue #32 for what the Space shows, and no doc says the Space still runs deploy 1, is not deployed, or has not been redeployed since August (#32)
Unverified signatureThe commit we deploy is also what origin/main is fast-forwarded to, and nothing can be pushed after the deploy without restarting the public Space, so a sentence that a deploy makes false would stay false. README: "Where things stand, 2026-11-17" says the README describes deploy 2's code, which replaces deploy 1 on the Space, points to SPACE_MODE_MATRIX 10.1-10.3 and the release notes, and to issue #32 for deploy 2's own check on the Space. The cold-start time points to #32 instead of "has not been measured"; the longest accepted reservation is "not measured by these runs"; the browser runs of 2026-11-16 "did not finish in that session" instead of "have not been run ... yet". docs/RELEASE_NOTES.md: "Deploy 2 - 2026-11-17", the code of this commit's tree, the Space result on #32 (also under "On the Space"); deploy 1 "ran" as 0cd54f2 until deploy 2, and main "was fast-forwarded" to 1e9fd20 that day. The other docs, the same way: SPACE_MODE_MATRIX's header and 10 intro, 10.2 and 10.3; …abc9150
4 days ago
-
docs: the packed size is quoted as the platform's progress bar printed it, 40.6G
Unverified signatureThe status boxes added in 8e7494f wrote "40.6 GB". The ZeroGPU packing bar on the 2026-11-13 boot printed "40.6G" and does not say whether that is GB or GiB, so the boxes now quote it as printed rather than assert a unit.e6561d9 -
docs: each August record opens with what has been superseded since, and where the current truth is
Unverified signatureThe seven historical documents keep their bodies. Each gets a short "Status on 2026-11-16" box at the top naming the passages that later measurements overtook, with the corrected figure and its source: - the stage-2 "ceiling between 25 and 49 frames" was the 47.375 GiB `large` partition; 704x480x49 passes on `xlarge` (MIG_ALLOCATOR_SIGHTINGS.md §2); - "≤ ~675.5 GPU-s" for UTC 2026-08-25: the persistent ledger reads ≤ 1528.672 over 20 runner calls, ~1622 with the hand-bounded manual calls; - "82.2 GB of 150 [OK]" is a reading taken before ZeroGPU packs; the platform packed 40.6 GB on the 2026-11-13 boot and the disk peak was never measured; - TEST_PLAN.md and ZEROGPU_DISK.md never got the 2026-08-25 retraction of "all eleven modes generate, content-verified"; - deploying is a merge commit on the Space head pushed as a fast-forward, not a push of the work commit; the download is at startup, 13 min 59 s cold; - hinode is retired and its scripts are gone; `auto` is no longer the…8e7494f
-
docs: every Space generation result was garbage — the invalidation, stated at the top
Unverified signatureThe deployed Space produced colour static for the entire project. Blanket int8 quantised the LTX-2.5 conditioning path (the eight LTX2AdaLayerNormSingle modulation heads plus the in/out projections), the denoiser stopped being steered, and the output became a correctly-shaped sample from the wrong distribution. Fixed at eb1aa9f, deployed e3b5333, verified by opening a frame. SPACE_MODE_MATRIX and SPACE_SESSION_2026-08-25 get a dated invalidation block at the very top, above existing content, per the operator's ruling. House rule honoured: nothing is rewritten, every measurement stands as history, the corrections are dated on top. Each block says the same six things: - Every generation result in the document was produced by a broken transformer. Space shas up to and including ff142d9 ran blanket int8. - MECHANICAL CLAIMS SURVIVE: all eleven modes execute; both fork-boundary directions carry their cargo; determinism and byte-reproducibility hold; tiled decode enables 1216x704x12…898dad2 -
docs: the audit's stale claims annotated — measurements stand, corrections dated on top
Unverified signatureA four-agent audit swept every doc for claims events had falsified (~30, with file:line). Fixed per the house rule: history is annotated, never rewritten — dated blockquote corrections point at current reality with evidence. Six follow-up-table rows get in-cell appends with status flips (the file's own precedent for rows); no finding is deleted, no captured record reworded. The delicate one first. SPACE_MODE_MATRIX §5's 'byte-identical to the result banked the previous day' was never verified: no day-1 hash was ever recorded, the match was file size (39,146 B) plus every decoded statistic to three decimals, and the day-1 file is gone with its host — so byte identity is now unverifiable against it. The §5 caveat ('Do NOT upgrade this') was correct and reports upgraded past it; the error ran in that direction, not the doc's. Corrected claim: identical in size and in all decoded statistics, byte identity UNVERIFIED; a same-seed re-run sha256-compared against the surviving 2026-08-22 arti…fd85c87
-
docs: 11 of 11 on the Space, and what that still does not claim
Unverified signatureThe two video-INPUT modes were re-run at Space sha 287ba9c carrying d2eb6f2 and both pass, so follow-up #1 closes: 19.95s / 63,438 B / -19.5 dBFS for video_to_audio, 13.67s / 33,812 B / 9 distinct frames / mean |d| 17.277 for video_to_video. Both land inside the 13.67-21.59s band every successful generation occupies, against the 3.53-9.13s the failures took. The uint8 choice is recorded as the part that was load-bearing on the metered side rather than locally: upstream re-compresses a single-frame condition at the model CRF and rejects a non-uint8 array while doing it, so a float payload would have passed the offline gate and failed only on the Space. Fixture provenance is recorded with the result, because without it the numbers prove much less. The runner chains fixtures forward, and by the end of the matrix video.mp4 had drifted to audio_to_video's near-static 7,619 B output at mean |d| 0.905. Conditioning on that would have been unreadable in either direction. The fixture was rebu…17b589c -
docs: the fork-boundary fix is a pair, and half of it is a regression
Unverified signatureThree corrections to SPACE_MODE_MATRIX.md, plus four stale claims elsewhere. The fix for the fork boundary was recorded as one method: a `__reduce__` on LTXStudioError. Reading `spaces` rather than trying it shows that is wrong in both directions, and neither half works alone. `__reduce__` alone does nothing. ExceptionResult (types.py:28-31) carries only `traceback`, `error_cls` and `gradio_error`, and `exception_result` populates the third only when `isinstance(exc, gr.Error)`. Our exception object is therefore never transmitted at all, nothing is ever pickled, and `__reduce__` is never consulted. That also settles the open question of whether restoring picklability would restore the message: it would not, and no live run is needed to know it. A `gr.Error` subclass alone is worse than the status quo. It makes the object eligible for `gradio_error`, which must then cross a multiprocessing.Queue that our classes cannot pickle onto. `spaces` anticipates this at wrappers.py:272-275 and…f927c4e
-
docs: the Space generates, and the fix is not attributed
Unverified signatureThe retry at 8fb292d6 succeeded. A real 39,146-byte video, verified for content and not only for structure: pixel std 72.657, 9 of 9 frames distinct, mean frame-to-frame delta 18.894. Structure alone was not accepted because H9 established on this build that a green stream table and a silent file are indistinguishable, so the video got the analogue of the RMS check that caught it. The device-API contrast is the cleanest evidence in the file: fail=true twice against fail=false once, from the platform's own accounting. 02fe943 stays. The revert condition was "if the retry fails" and it did not fail, so the condition is discharged rather than overridden. Recorded as kept because it did not fail, NOT because it was shown to help. The caveat is given the same prominence as the success, in the banner and in its own section. The failing and succeeding runs differ by two things: 02fe943, and a full container rebuild. That is not the clean A/B we had for the RSS number, where an AST diff show…1bcc98e -
docs: the inherited ceiling is falsified and the RAM arithmetic is dead
Unverified signatureThe route-B run at 8fb292d6 reached 87.13 GiB = 93.56 GB peak RSS and was not killed. Three findings, none of which depend on how the retry lands. 1. The ~68 GB ceiling is FALSIFIED, not merely unconfirmed. A process on this platform survived 25.56 GB above it. Replaced with our own measured lower bound, above 93.56 GB. Every occurrence audited: the two inside quoted historical claims are kept byte-identical and marked superseded, the projection table's column header goes from a warning to falsified so it cannot be read as merely unverified, and the rest sit inside the falsification. Consequence recorded: headroom at the 60.0 GB run was at least 33.56 GB, not the ~8 GB assumed, so the cgroup-OOM hypothesis was built inside a gap that did not exist. 2. ZEROGPU_DISK.md 4.2's single mechanism has now decided a third question, against us: quantising the text encoder materialised what safetensors kept lazy, and peak RSS rose 31.24 GiB in a change made to lower it. Section 9 already said s…fff04a9 -
docs: annotate the startup log quote, do not rewrite it
Unverified signatureThe quoted line at section 5 reads `loaded transformer as int8` and the code no longer prints that. The quote stays byte-identical: it is a record of what that run printed, and editing a captured measurement to match today's source would falsify it. Pure addition -- 9 lines, 0 removed. The annotation does NOT say the line under-reported. It did not. At the time of that run `components_to_quantize=["transformer"]` and exactly one component was quantised, so `loaded transformer as int8` was accurate. The constant that now names two did not exist until 02fe943, which is later than this document; the single bitsandbytes module line in the captured log corroborates the count independently. Records the consequence that actually matters for the next decision: the 55.89 GiB peak RSS in section 6 was measured with ONE component in int8, not two. The ceiling inference is unaffected -- the process was not killed at 60.0 GB, so the ceiling is above it either way -- but anyone reading that figure…1802bfd -
docs: the recommendation in this file was a false instruction
Unverified signatureSection 9 told the reader to flip LTX25_TEXT_ENCODER_PRECISION to int8 and read the peak-RSS line. That variable is wired to nothing. Verified by mutation rather than by reading, and independently of the report that raised it: the name appears in no source file; load_config() returns an identical RuntimeConfig with the variable unset and set to 'int8'; an illegal value is accepted silently. The contrast is LTX25_QUANTIZATION, which raises ConfigError on the same nonsense input. It is not half-wired. It is not a RuntimeConfig field at all, so it is never resolved, validated or logged. PLAN.md records that both precision variables were REMOVED because an advertised lever attached to nothing is worse than no lever -- the Space Variables were leftovers from before that removal. And the running Space announced it, in a line this document already quoted in section 5: "dtype : bf16 (fixed; no precision switch)". Nobody read it. The evidence was in the log everyone was working from. What i…b56acd4 -
docs: the Space starts and cannot generate
Unverified signatureFirst generation ever attempted on the live private Space. Startup succeeded, the generation failed: the forked ZeroGPU worker was killed twice with no Python traceback and spaces returned 'GPU task aborted'. This file is the only place these measurements exist. Organised around keeping what we MEASURED apart from what we INHERITED. The RAM ceiling has two numbers and only one is ours: startup reached 55.89 GiB = 60.0 GB and was not killed, so the ceiling is measurably above 60.0 GB -- our own datum, the first real information anyone here has had about that number. The ~68 GB figure remains a community cgroup observation we have never measured, and every headroom computed against it is marked as inheriting that status, including the table header that states the assumption. The absence of a traceback is recorded as the finding rather than as a gap in the report. An exception unwinds and leaves a stack; a SIGKILL leaves nothing. That absence is what most constrains the hypotheses -- an…51e1985