Dontopedia
Explore

Iteration 90

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

Iteration 90 has 32 facts recorded in Dontopedia across 4 references, with 2 live disagreements.

32 facts·26 predicates·4 sources·2 in dispute

Mostly:iteration number(3), rdf:type(2), loss value(2)

Maturity scale raw canonical shape-checked rule-derived certified

Rdf:typein disputerdf:type

Loss Valuein disputelossValue

  • nan[4]all time · 111
  • 3.1995[2]sourceall time · 114

Has Learning RatehasLearningRate

  • 0.0000381[3]sourceall time · 115
  • 0.0000381[1]all time · Part 114

Belongs tobelongsTo

Has Gpu Cache MemoryhasGpuCacheMemory

  • 34.8GB[1]all time · Part 114

Has Gpu MemoryhasGpuMemory

  • 1.0GB[1]all time · Part 114

Has Gpu Peak MemoryhasGpuPeakMemory

  • 34.3GB[1]all time · Part 114

Has Iteration NumberhasIterationNumber

  • 90[1]all time · Part 114

Has LosshasLoss

  • 3.1995[1]all time · Part 114

Has Process MemoryhasProcessMemory

  • 3.2GB[1]all time · Part 114

Has Tokens Per SecondhasTokensPerSecond

  • 6766[1]all time · Part 114

Temporally PrecedestemporallyPrecedes

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.

consistsOfConsists of(1)

temporallyPrecedesTemporally Precedes(1)

Other facts (17)

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.

17 facts
PredicateValueRef
Iteration Number90[3]
Iteration Number90[2]
Iteration Number90[4]
Learning Rate0.0000381[2]
Learning Rate0.0000381[4]
Process Memory Usage3.2[3]
Gpu Cache Memory11.8[3]
Gpu Memory Peak12[3]
Gpu Memory Usage1[3]
Has Token Throughput17631[3]
Has Loss Value3.1631[3]
Part of RunTraining Run Grouped1[3]
Process Memory Gb3.2[2]
Gpu Cache Gb34.8[2]
Gpu Memory Peak Gb34.3[2]
Gpu Memory Usage Gb1[2]
Tokens Per Second6766[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.

belongsToblah/watt-activation/part-114
ex:training-run-linear-seq2048
gpuCacheGBblah/watt-activation/114
34.8
gpuCacheMemoryblah/watt-activation/115
11.8
gpuMemoryPeakblah/watt-activation/115
12
gpuMemoryPeakGBblah/watt-activation/114
34.3
gpuMemoryUsageblah/watt-activation/115
1
gpuMemoryUsageGBblah/watt-activation/114
1
hasGpuCacheMemoryblah/watt-activation/part-114
34.8GB
hasGpuMemoryblah/watt-activation/part-114
1.0GB
hasGpuPeakMemoryblah/watt-activation/part-114
34.3GB
hasIterationNumberblah/watt-activation/part-114
90
hasLearningRateblah/watt-activation/115
0.0000381
hasLearningRateblah/watt-activation/part-114
0.0000381
hasLossblah/watt-activation/part-114
3.1995
hasLossValueblah/watt-activation/115
3.1631
hasProcessMemoryblah/watt-activation/part-114
3.2GB
hasTokensPerSecondblah/watt-activation/part-114
6766
hasTokenThroughputblah/watt-activation/115
17631
iterationNumberblah/watt-activation/115
90
iterationNumberblah/watt-activation/114
90
iterationNumberblah/watt-activation/111
90
learningRateblah/watt-activation/114
0.0000381
learningRateblah/watt-activation/111
0.0000381
lossValueblah/watt-activation/111
nan
lossValueblah/watt-activation/114
3.1995
partOfRunblah/watt-activation/115
ex:training-run-grouped1
processMemoryGBblah/watt-activation/114
3.2
processMemoryUsageblah/watt-activation/115
3.2
typeblah/watt-activation/114
ex:Iteration
typeblah/watt-activation/115
ex:Iteration
temporallyPrecedesblah/watt-activation/part-114
ex:iteration-100
tokensPerSecondblah/watt-activation/114
6766

References (4)

4 references
  1. [1]Part 11410 facts
    customctx:discord/blah/watt-activation/part-114
  2. customctx:discord/blah/watt-activation/114
    • full textwatt-activation-114
      text/plain2 KBdoc:agent/watt-activation-114/bc8b5ab5-bbc4-4aea-ac89-9211a96127c3
      Show excerpt
      [2026-03-08 22:42] xenonfun: Now add val evaluation in the training loop, after the checkpoint save block [2026-03-08 22:44] xenonfun: ``` === linear (seq=2048) === Training HarmonicGPT | attn=linear d=768 L=12 H=12 seq=2048 Parameters: 86,
  3. customctx:discord/blah/watt-activation/115
    • full textwatt-activation-115
      text/plain3 KBdoc:agent/watt-activation-115/105443fa-23c3-42fc-a3a6-065e0c10a484
      Show excerpt
      [2026-03-08 22:44] xenonfun: ``` === grouped1 (seq=2048) === Training HarmonicGPT | attn=grouped1 d=768 L=12 H=12 seq=2048 Parameters: 86,825,760 Loaded 5,000 train sequences of length 2048 iter 10/100 | loss 5.4746 | lr 3.00e-04 | tok/
  4. [4]1113 facts
    customctx:discord/blah/watt-activation/111

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.