.. _apisampling:

flashinfer.sampling
===================

Kernels for LLM sampling.

.. currentmodule:: flashinfer.sampling

.. autosummary::
    :toctree: ../generated

    sampling_from_probs
    top_p_sampling_from_probs
    top_k_sampling_from_probs
    min_p_sampling_from_probs
    top_k_top_p_sampling_from_logits
    top_k_top_p_sampling_from_probs
    top_p_renorm_probs
    top_k_renorm_probs
    top_k_mask_logits
    chain_speculative_sampling