25M training run
From Dontopedia, the open, paraconsistent wiki. (Last updated 2026-06-06.)
25M training run has 24 facts recorded in Dontopedia across 2 references, with 3 live disagreements.
Mostly:has checkpoint at step(5), has batch size(2), has estimated duration(2)
Maturity scale
raw canonical shape-checked rule-derived certifiedInbound 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.
announcesProgressAnnounces Progress(1)
- Xenonfun
ex:xenonfun
framesTrainingAsOngoingFrames Training As Ongoing(1)
- Message 01 50
ex:message-01-50
partOfPart of(1)
- Step 1000
ex:step-1000
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.
| Predicate | Value | Ref |
|---|---|---|
| Has Checkpoint at Step | 2000 | [2] |
| Has Checkpoint at Step | 4000 | [2] |
| Has Checkpoint at Step | 6000 | [2] |
| Has Checkpoint at Step | 8000 | [2] |
| Has Checkpoint at Step | 10000 | [2] |
| Has Batch Size | 16 | [1] |
| Has Batch Size | 16 | [2] |
| Has Estimated Duration | 95 minutes | [1] |
| Has Estimated Duration | ~95 minutes | [2] |
| Is From Scratch | 25m Model | [1] |
| Is From Scratch | true | [2] |
| Estimated to Complete in | 95 minutes | [1] |
| Has Expected Tok Per Sec | 7000 | [1] |
| Has Seq Len | 256 | [1] |
| Has Steps | 10000 | [1] |
| Kicked Off After | Training Run 50m | [1] |
| Processes Total Tokens | 40000000 | [1] |
| Saves Checkpoints at | steps 2K, 4K, 6K, 8K, 10K | [1] |
| Has Model Size | 25M | [2] |
| Has Step Count | 10000 | [2] |
| Has Sequence Length | 256 | [2] |
| Has Total Token Count | 40000000 | [2] |
| Has Estimated Speed | 7000 | [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)
ctx:discord/blah/watt-activation/part-644ctx:discord/blah/watt-activation/641- full textwatt-activation-641text/plain2 KB
doc:agent/watt-activation-641/f90a56d8-d0f5-46e7-92ee-3c668e2366e4Show excerpt
[2026-04-16 01:42] xenonfun: ```bash ➜ ~/MS/HarmonicRust git:(master) ✗ ./target/release/wave_native_train --size 50m --data /tmp/domain_tinystories.bin --mode full --seq-len 256 --batch-size 32 --total-steps 10 --lr 0.001 --warmup 0 --log…
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.