Dontopedia

Training Run Linear Seq2048

From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-05-02.)

Training Run Linear Seq2048 has 24 facts recorded in Dontopedia across 1 reference, with 1 live disagreement.

24 facts·15 predicates·1 sources·1 in dispute

Mostly:consists of(10), achieves low loss(1), commits to success(1)

Maturity scale raw canonical shape-checked rule-derived certified

Consists ofin disputeconsistsOf

Inbound mentions (14)

Other subjects in dontopedia point AT this entity as a value. These are inverse relationships — e.g. "X motherOf this subject" — and answer questions the forward facts can't. Grouped by predicate.

belongsToBelongs to(11)

expressesPositiveStanceTowardExpresses Positive Stance Toward(1)

performsDemonstrationPerforms Demonstration(1)

presentsPresents(1)

Other facts (14)

The long tail: predicates that appear too rarely to warrant their own section. Filter or scroll to find a specific one. Each row links to its source.

14 facts
PredicateValueRef
Achieves Low Loss3.1939[1]
Commits to Successtraining complete[1]
Exhibits Learning Rate DecayLearning Rate Scheduler[1]
Has Stable Gpu Usage1.0GB active[1]
Has Statuscomplete[1]
Has Total Parameters86825760[1]
Labeled Aslinear (seq=2048)[1]
Loaded Sequence Count5000[1]
Loaded Sequence Length2048[1]
Presupposes Hardware AvailabilityGpu[1]
References ML Training ParadigmSupervised Sequence Modeling[1]
Reports With CertaintyLoss Metrics[1]
Shows Tokens Per Second Stabilizationaround 6760[1]
Trains ModelHarmonic Gpt[1]

Timeline

Timeline axis is valid_time — when each source says the fact was true in the world, not when Dontopedia learned about it. Retracted rows are kept for provenance; coloured stripes indicate the context kind.

achievesLowLossblah/watt-activation/part-114
3.1939
commitsToSuccessblah/watt-activation/part-114
training complete
consistsOfblah/watt-activation/part-114
ex:iteration-10
consistsOfblah/watt-activation/part-114
ex:iteration-100
consistsOfblah/watt-activation/part-114
ex:iteration-20
consistsOfblah/watt-activation/part-114
ex:iteration-30
consistsOfblah/watt-activation/part-114
ex:iteration-40
consistsOfblah/watt-activation/part-114
ex:iteration-50
consistsOfblah/watt-activation/part-114
ex:iteration-60
consistsOfblah/watt-activation/part-114
ex:iteration-70
consistsOfblah/watt-activation/part-114
ex:iteration-80
consistsOfblah/watt-activation/part-114
ex:iteration-90
exhibitsLearningRateDecayblah/watt-activation/part-114
ex:learning-rate-scheduler
hasStableGpuUsageblah/watt-activation/part-114
1.0GB active
hasStatusblah/watt-activation/part-114
complete
hasTotalParametersblah/watt-activation/part-114
86825760
labeledAsblah/watt-activation/part-114
linear (seq=2048)
loadedSequenceCountblah/watt-activation/part-114
5000
loadedSequenceLengthblah/watt-activation/part-114
2048
presupposesHardwareAvailabilityblah/watt-activation/part-114
ex:gpu
referencesMlTrainingParadigmblah/watt-activation/part-114
ex:supervised-sequence-modeling
reportsWithCertaintyblah/watt-activation/part-114
ex:loss-metrics
showsTokensPerSecondStabilizationblah/watt-activation/part-114
around 6760
trainsModelblah/watt-activation/part-114
ex:harmonic-gpt

References (1)

1 references
  1. [1]Part 11424 facts
    ctx:discord/blah/watt-activation/part-114

See also

Keep researching

Missing something or suspicious of what's here? Kick off a research session — a Claude agent will investigate, cite its sources, and file new facts into a dedicated context you can review before accepting into the shared view.