liblloyal 1.0.0
Branched Inference for llama.cpp
Loading...
Searching...
No Matches
lloyal::branch::ForkOpts Struct Reference

Options controlling what state fork() clones from source to child. More...

#include <lloyal/branch.hpp>

Public Attributes

bool clone_logits = true
 If true (default), copy src->logits_snapshot to child (~n_vocab*4 bytes, ~600KB for 150k-vocab models).
 

Detailed Description

Options controlling what state fork() clones from source to child.

Default {} preserves fork's full "sample same distribution" contract. Set clone_logits=false for prefill-overwrite consumers (rerank leaves, embedding probes) that fork → prefill → read → prune and never sample the parent's distribution.

Definition at line 1352 of file branch.hpp.

Member Data Documentation

◆ clone_logits

bool lloyal::branch::ForkOpts::clone_logits = true

If true (default), copy src->logits_snapshot to child (~n_vocab*4 bytes, ~600KB for 150k-vocab models).

If false, leave child has_logits=false until subsequent prefill()/step(); calling sample() on the child before that returns -1 (the kernel's no-logits sentinel; the SDK Branch.sample() converts this to BranchSampleError).

Examples
/home/runner/work/liblloyal/liblloyal/include/lloyal/branch.hpp.

Definition at line 1358 of file branch.hpp.


The documentation for this struct was generated from the following file: