Training Run 2026 05 01
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
Training Run 2026 05 01 has 100 facts recorded in Dontopedia across 2 references, with 6 live disagreements.
Mostly:has elapsed time at step(16), has tokens per second at step(16), has loss at step(16)
Maturity scale
raw canonical shape-checked rule-derived certifiedHas Elapsed Time at Stepin disputehasElapsedTimeAtStep
Has Tokens Per Second at Stepin disputehasTokensPerSecondAtStep
Has Loss at Stepin disputehasLossAtStep
Has Bpb at Stepin disputehasBpbAtStep
Recorded at Stepin disputerecordedAtStep
Has Iterations Per Second at Stepin disputehasIterationsPerSecondAtStep
Last Ten BpblastTenBPB
- 1.2241[1]all time · 700
Final Tokens Per SecondfinalTokensPerSecond
- 12860[1]sourceall time · 700
Best Logged Bpb StepbestLoggedBPBStep
- 21100[1]sourceall time · 700
Best Logged BpbbestLoggedBPB
- 1.0219[1]sourceall time · 700
Has Final LosshasFinalLoss
- 0.8237[1]sourceall time · 700
Has Final BpbhasFinalBPB
- 1.1883[1]sourceall time · 700
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.
comparesEntityCompares Entity(3)
- Comparison Active Params
ex:comparison-active-params - Comparison Parameters
ex:comparison-parameters - Comparison Wall Time
ex:comparison-wall-time
Other facts (7)
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 |
|---|---|---|
| Has Wall Time Duration | 5.2h | [2] |
| Has Capability Condition | when data arrives | [2] |
| Has Context Length Capability | 4K | [2] |
| Has Active Parameters Per Token | 40000000 | [2] |
| Has Total Parameters | 125000000 | [2] |
| Has Log Entry | Log Entry 2026 05 01 03 28 | [2] |
| Rdf:type | Model Training Run | [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.
References (2)
- custom
ctx:discord/blah/watt-activation/700- full textwatt-activation-700text/plain3 KB
doc:agent/watt-activation-700/35753342-3814-4c1d-a934-fd039048c39eShow excerpt
[2026-05-01 08:24] xenonfun: I actually thought we wanted to feed the whole sliced sequences not random 2K chops. but its working okay still think this it wrong. (files: Screenshot_2026-05-01_at_4.23.45_AM.png) [2026-05-01 08:50] xenonfun: …
- custom
ctx:discord/blah/watt-activation/697- full textwatt-activation-697text/plain3 KB
doc:agent/watt-activation-697/0bcf6f97-e39d-494c-8129-e73c0ac78f73Show excerpt
[2026-05-01 03:28] xenonfun: ``` step | bpb | loss | tok/s | iter/s | elapsed -------+--------+---------+------------+---------+---------- 1 | 5.809 | 4.0263 | 576 | 0.04 | 0m28s 50 | 4.926 | …
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.