For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /max/get-started.md).

Mojo function

topp_minp_sampling_kernel

def topp_minp_sampling_kernel[dtype: DType, out_idx_type: DType, is_top_p: Bool](p_thresholds_: UnsafePointer[Scalar[dtype], MutUntrackedOrigin], sorted_probs_: UnsafePointer[Scalar[dtype], MutUntrackedOrigin], sorted_ids_: UnsafePointer[Scalar[out_idx_type], MutUntrackedOrigin], out_token_ids: UnsafePointer[Scalar[out_idx_type], MutUntrackedOrigin], skip_sort: UnsafePointer[Scalar[DType.bool], MutUntrackedOrigin], vocab_size: Int)

Top P-Min P sampling kernel.

Parameters:

dtype (DType): DType - scalar values dtype.
out_idx_type (DType): DType - output index type.
is_top_p (Bool): Bool - Whether to use Top-P (True) or Min-P (False) sampling.