New Training Run Novels
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
New Training Run Novels has 37 facts recorded in Dontopedia across 1 reference, with 1 live disagreement.
Mostly:tracking enabled(4), uses monitoring method(1), has batch sizing strategy(1)
Maturity scale
raw canonical shape-checked rule-derived certifiedTracking Enabledin disputetrackingEnabled
- Candidates[1]sourceall time · 25
- Mi Profiles[1]sourceall time · 25
- Reports[1]sourceall time · 25
- Summaries[1]sourceall time · 25
Uses Monitoring MethodusesMonitoringMethod
- Cusum Monitoring[1]sourceall time · 25
Has Batch Sizing StrategyhasBatchSizingStrategy
- Adaptive Batch Sizing[1]sourceall time · 25
Selection StrategyselectionStrategy
- Top K by Val Loss[1]sourceall time · 25
Diversity Bonus SchedulediversityBonusSchedule
- Cosine Decay[1]sourceall time · 25
Diversity BonusdiversityBonus
- 0.08[1]sourceall time · 25
Mutation Rate GoalmutationRateGoal
- Exploration[1]sourceall time · 25
Mutation Rate DescriptionmutationRateDescription
- aggressive[1]sourceall time · 25
Mutation RatemutationRate
- 0.6[1]sourceall time · 25
Total Search StepstotalSearchSteps
- 48000[1]sourceall time · 25
Generation CountgenerationCount
- 16[1]sourceall time · 25
Population SizepopulationSize
- 6[1]sourceall time · 25
Inbound mentions (1)
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.
announcedEventAnnounced Event(1)
- Omega Bot
ex:omega-bot
Other facts (22)
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.
| Predicate | Value | Ref |
|---|---|---|
| Search Steps Per Candidate | 500 | [1] |
| Symbiogenesis Config Includes | Activation Variety | [1] |
| Uses Tokenizer | Bpe Tokenizer | [1] |
| Optimizer | Adamw | [1] |
| Learning Rate | 0.0003 | [1] |
| Batch Size | 16 | [1] |
| Context Block Size | 512 | [1] |
| Model Layer Count | 8 | [1] |
| Model Attention Heads | 8 | [1] |
| Model Embedding Dimension | 384 | [1] |
| Model Size in Memory | 60.7 | [1] |
| Model Parameter Count | 15.9 | [1] |
| Precision Format | F16 | [1] |
| Uses Backend | Helios | [1] |
| Execution Region | US Central1 B | [1] |
| Executed on | G2 Standard 4 | [1] |
| Uses Hardware | Nvidia L4 | [1] |
| Domain | novels | [1] |
| Uses Dataset | Novels All Txt | [1] |
| Has Run Id | novels_all_20260225152704_30v8 | [1] |
| Rdfs:label | 50k Symbio Novels Training Run | [1] |
| Rdf:type | Training Run | [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.
References (1)
- custom
ctx:discord/blah/training-and-evals/25- full texttraining-and-evals-25text/plain2 KB
doc:agent/training-and-evals-25/e24e7b5b-4c3d-43fe-b0a2-7c1baa240994Show excerpt
[2026-02-25 14:57] omega [bot]: **New: 3D Radial Training Visualization (Three.js)** Added an interactive 3D scene to the run detail page below the oscillation chart. Concentric rings map training metrics radially around the timeline: - *…
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.