For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).
Mojo module
reductions
Row-based row-wise reductions (free-form body layer).
Pure-reduction entry points (reduce_sum / reduce_max / reduce_min /
reduce_product / reduce_mean / reduce_argmax / reduce_argmin /
reduce_min_and_max) built on the algorithm.rowwise scaffolder and the
algorithm.reduce_op monoid library. Each authors one free-form body β a
single reduce phase plus a per-row emit β and runs on both CPU and GPU.
Functionsβ
- β
reduce_argmax: - β
reduce_argmin: Finds the argmin alongreduce_dimvia the Row layer: oneArgMinphase (native dtype, lower index wins ties) plus a per-rowemitof the int64 index. - β
reduce_max: - β
reduce_mean: - β
reduce_min: - β
reduce_min_and_max: Computes min and max in one axis walk (MinMaxmonoid), two outputs. - β
reduce_product: - β
reduce_sum: Sums input values alongreduce_dimvia the Row free-form body layer: oneReduceSumphase plus a per-rowemit.