implementation detail · filed under model architecture
Pooled-global-batch quantile estimation
Estimates a quantile from the pooled global batch rather than averaging per-rank quantiles.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
estimate is the quantile of the pooled global batch rather than an average of per-rank quantiles, which generally differs.
usedmodel architecturein Kimi K3Moonshot AI
Filed alongside
Other methods under model architecture :: channel mixer :: mixture of experts :: expert load balancing.
Auxiliary-loss-free load balancingQuantile BalancingHistogram-based quantile estimationAuxiliary-loss load balancingExact coordinate minimizationExpert bias update factorExpert Parallelism Load BalancingExponential moving average of estimated quantilesformHCLoad balancingMaxVioModality-specific auxiliary-loss-free load balancingPersistent load balancingRound-robin load balancing