layer l's heads' attention as one streaming launch (Realization.Nvidia.
SM86.StreamingAttentionSM86): a block per 64 queries
of a head, the output written into the layer's merged half, each row's
base-2 log-sum-exp into its log-sum-exp (the backward's)
2025def cgStreamingAttention = (lambda unrestricted l : Nat .
2026 (cgLaunch
2027 cgImageAttentionForward
2028 (naturalDivideUnchecked cgSeq 64)
2029 cgHeads
2030 1
2031 (cgPointer
2032 (cgArgument 0)
2033 (cgQueriesOf l)
2034 (cgPointer
2035 (cgArgument 1)
2036 (cgKeysOf l)
2037 (cgPointer
2038 (cgArgument 2)
2039 (cgValuesTOf l)
2040 (cgPointer
2041 (cgArgument 3)
2042 (cgMergedHalfOf l)
2043 (cgPointer (cgArgument 4) (cgLogSumExpOf l) (cgSlot (cgArgument 5) cgStreamingScale cgNoSlots))))))))The compiler supplied declaration spans and resolved links from this source snapshot. This page does not assert that this file belongs to a checked closure.