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.
Mostly:consists of(10), achieves low loss(1), commits to success(1)
Maturity scale
raw canonical shape-checked rule-derived certifiedConsists ofin disputeconsistsOf
- Iteration 10[1]all time · Part 114
- Iteration 100[1]all time · Part 114
- Iteration 20[1]all time · Part 114
- Iteration 30[1]all time · Part 114
- Iteration 40[1]all time · Part 114
- Iteration 50[1]all time · Part 114
- Iteration 60[1]all time · Part 114
- Iteration 70[1]all time · Part 114
- Iteration 80[1]all time · Part 114
- Iteration 90[1]all time · Part 114
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)
- Final Training Stats
ex:final-training-stats - Iteration 10
ex:iteration-10 - Iteration 100
ex:iteration-100 - Iteration 20
ex:iteration-20 - Iteration 30
ex:iteration-30 - Iteration 40
ex:iteration-40 - Iteration 50
ex:iteration-50 - Iteration 60
ex:iteration-60 - Iteration 70
ex:iteration-70 - Iteration 80
ex:iteration-80 - Iteration 90
ex:iteration-90
expressesPositiveStanceTowardExpresses Positive Stance Toward(1)
- Xenonfun
ex:xenonfun
performsDemonstrationPerforms Demonstration(1)
- Chat Message 2026 03 08 22 44
ex:chat-message-2026-03-08-22-44
presentsPresents(1)
- Chat Message 2026 03 08 22 44
ex:chat-message-2026-03-08-22-44
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.
| Predicate | Value | Ref |
|---|---|---|
| Achieves Low Loss | 3.1939 | [1] |
| Commits to Success | training complete | [1] |
| Exhibits Learning Rate Decay | Learning Rate Scheduler | [1] |
| Has Stable Gpu Usage | 1.0GB active | [1] |
| Has Status | complete | [1] |
| Has Total Parameters | 86825760 | [1] |
| Labeled As | linear (seq=2048) | [1] |
| Loaded Sequence Count | 5000 | [1] |
| Loaded Sequence Length | 2048 | [1] |
| Presupposes Hardware Availability | Gpu | [1] |
| References ML Training Paradigm | Supervised Sequence Modeling | [1] |
| Reports With Certainty | Loss Metrics | [1] |
| Shows Tokens Per Second Stabilization | around 6760 | [1] |
| Trains Model | Harmonic 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.
References (1)
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.