Companion

Functions

Link copied to clipboard
fun dense(numLayers: Int, numHeads: Int, headDim: Int, maxSeqLen: Int): KvCacheStore

Create an uncompressed FP32 KV cache (baseline).

Link copied to clipboard

Create a KV cache from a TurboQuantPreset.

Link copied to clipboard
fun turboQuant(preset: String, numLayers: Int, numHeads: Int, headDim: Int, maxSeqLen: Int): KvCacheStore

Create a TurboQuant-compressed KV cache from a named preset.

fun turboQuant(numLayers: Int, numHeads: Int, headDim: Int, maxSeqLen: Int, keyBits: Int = 4, valueBits: Int = 4, useQjl: Boolean = false): KvCacheStore

Create a TurboQuant-compressed KV cache with custom bit budgets.