Dontopedia

Current Training

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

Current Training has 20 facts recorded in Dontopedia across 10 references.

20 facts·20 predicates·10 sources

Mostly:has more data than before(1), has one full epoch(1), has tokens(1)

Maturity scale raw canonical shape-checked rule-derived certified

Inbound mentions (3)

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.

plannedAfterPlanned After(1)

showsWorseningMomentarilyShows Worsening Momentarily(1)

stallsOccasionallyStalls Occasionally(1)

Other facts (20)

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.

20 facts
PredicateValueRef
Has More Data Than Before2.6×[1]
Has One Full Epochtrue[1]
Has Tokens55.7M[1]
Is Big Runtrue[1]
Uses Lower Lr Warmuptrue[2]
Expected to Convergetrue[2]
Will ImproveBest Score[2]
Has Not Beaten BestFt1[2]
Compared toPrerestart[3]
Out of Warmup Phasetrue[3]
Ppl Change+3 ppl @ 42K[3]
Lacks Prefetchertrue[4]
Second ExperimentBranching From Base[5]
Uses DatasetsFineweb Plus Tinystories[6]
Insufficient forcoherent instruction-following or factual accuracy[6]
Has Limited Context Remainingnull[7]
Nearing Completionnull[7]
UsesOld Code Pre Givens[8]
Has Tok Per S10700[9]
Is Worse at MomentLlrd 0.8[10]

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.

20 undated
—
hasMoreDataThanBeforeblah/random/part-27
2.6×
—
hasOneFullEpochblah/random/part-27
true
—
hasTokensblah/random/part-27
55.7M
—
isBigRunblah/random/part-27
true
—
usesLowerLrWarmupblah/watt-activation/part-34
true
—
expectedToConvergeblah/watt-activation/part-34
true
—
willImproveblah/watt-activation/part-34
ex:best-score
—
hasNotBeatenBestblah/watt-activation/part-34
ex:ft1
—
comparedToblah/watt-activation/part-99
ex:prerestart
—
outOfWarmupPhaseblah/watt-activation/part-99
true
—
pplChangeblah/watt-activation/part-99
+3 ppl @ 42K
—
lacksPrefetcherblah/watt-activation/part-126
true
—
secondExperimentblah/watt-activation/part-144
ex:branching-from-base
—
usesDatasetsblah/watt-activation/part-166
ex:fineweb-plus-tinystories
—
insufficientForblah/watt-activation/part-166
coherent instruction-following or factual accuracy
—
hasLimitedContextRemainingblah/watt-activation/part-217
null
—
nearingCompletionblah/watt-activation/part-217
null
—
usesblah/watt-activation/part-397
ex:old-code-pre-givens
—
hasTokPerSblah/watt-activation/part-658
10700
—
isWorseAtMomentblah/watt-activation/part-37
ex:llrd-0.8

References (10)

10 references
  1. [1]Part 274 facts
    ctx:discord/blah/random/part-27
  2. [2]Part 344 facts
    ctx:discord/blah/watt-activation/part-34
  3. [3]Part 993 facts
    ctx:discord/blah/watt-activation/part-99
  4. [4]Part 1261 fact
    ctx:discord/blah/watt-activation/part-126
  5. [5]Part 1441 fact
    ctx:discord/blah/watt-activation/part-144
  6. [6]Part 1662 facts
    ctx:discord/blah/watt-activation/part-166
  7. [7]Part 2172 facts
    ctx:discord/blah/watt-activation/part-217
  8. [8]Part 3971 fact
    ctx:discord/blah/watt-activation/part-397
  9. [9]Part 6581 fact
    ctx:discord/blah/watt-activation/part-658
  10. [10]Part 371 fact
    ctx:discord/blah/watt-activation/part-37

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.

↑