FA-99258 / Distributed training aggregation / Member archive
Per-worker loss means receive equal weight despite different token counts · case 03
Per-worker loss means receive equal weight despite different token counts.
Case contract
Rows are [loss_sum, valid_count]; return global token mean or None when total count is zero.
Why this case matters
A toy distributed-training contract makes replica and shard contribution rules explicit without claiming framework compatibility or performance.
One recorded failure
Sample boundary fixtureThis sample comes from the broken implementation of a controlled reproducer.
| Boundary fixture | Actual | Expected | Outcome |
|---|---|---|---|
| unequal tokens | 6.0 | 7.5 | Failed |
MEMBER ARCHIVE
The complete case is available to members.
This record includes three runnable implementations, regression fixtures, execution results, and source hashes.
Member access is invitation-based. Sign in with your invited account to inspect the sources.
Sign in to the archive ↗