Dontopedia

Iteration 62500

From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)

Iteration 62500 has 23 facts recorded in Dontopedia across 2 references, with 2 live disagreements.

23 facts·18 predicates·2 sources·2 in dispute

Mostly:iterations per second(2), best loss so far(2), loss(2)

Maturity scale raw canonical shape-checked rule-derived certified

Inbound mentions (2)

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.

comparesSubjectCompares Subject(1)

fasterThanFaster Than(1)

Other facts (23)

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.

23 facts
PredicateValueRef
Iterations Per Second17.5[1]
Iterations Per Second17.5[2]
Best Loss So Far3.2235[1]
Best Loss So Far3.2235[2]
Loss4.1618[1]
Loss4.1618[2]
Tokens Per Second4.5K[1]
Tokens Per Second4500[2]
Learning Rate1.30e-04[1]
Learning Rate0.00013[2]
Progress Percent62.50%[1]
Avg100 Loss4.2764[1]
Avg100 Ppl71.98[1]
Eta35m 42s[1]
Ppl64.19[1]
Rdf:typeTraining Iteration[2]
Iteration Number62500[2]
Perplexity64.19[2]
Average Loss Over1004.2764[2]
Average Perplexity Over10071.98[2]
Percent Complete62.5[2]
Estimated Time Remaining35m 42s[2]
Has Compilation Modenot compiled (auto fused kernels)[2]

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.

iterationsPerSecondblah/watt-activation/part-32
17.5
progressPercentblah/watt-activation/part-32
62.50%
avg100Lossblah/watt-activation/part-32
4.2764
avg100Pplblah/watt-activation/part-32
71.98
bestLossSoFarblah/watt-activation/part-32
3.2235
lossblah/watt-activation/part-32
4.1618
tokensPerSecondblah/watt-activation/part-32
4.5K
etablah/watt-activation/part-32
35m 42s
learningRateblah/watt-activation/part-32
1.30e-04
pplblah/watt-activation/part-32
64.19
typeblah/watt-activation/32
ex:TrainingIteration
iterationNumberblah/watt-activation/32
62500
lossblah/watt-activation/32
4.1618
perplexityblah/watt-activation/32
64.19
averageLossOver100blah/watt-activation/32
4.2764
averagePerplexityOver100blah/watt-activation/32
71.98
bestLossSoFarblah/watt-activation/32
3.2235
learningRateblah/watt-activation/32
0.00013
iterationsPerSecondblah/watt-activation/32
17.5
tokensPerSecondblah/watt-activation/32
4500
percentCompleteblah/watt-activation/32
62.5
estimatedTimeRemainingblah/watt-activation/32
35m 42s
hasCompilationModeblah/watt-activation/32
not compiled (auto fused kernels)

References (2)

2 references
  1. [1]Part 3210 facts
    ctx:discord/blah/watt-activation/part-32
  2. [2]3213 facts
    ctx:discord/blah/watt-activation/32
    • full textwatt-activation-32
      text/plain3 KBdoc:agent/watt-activation-32/8e8fa426-eb3a-4cc0-8c6a-550ba45f969f
      Show excerpt
      [2026-03-06 21:37] xenonfun: ====================================================================== Evaluation at iteration 20000 (~20% thru data) ====================================================================== Prompt: 'virtue is' Ge

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.