onGeneration method
Call on every generation with its conversation's sampling and whether
the caller set any of it. Returns null when there is nothing to say;
otherwise the message, with warn true for a release warning — once per
engine, when the caller chose either sampler — and false for the verbose
log, when both are defaults (a Qwen session with thinking on after one
with it off).
Implementation
({String message, bool warn})? onGeneration(
Sampler sampling, {
required bool explicit,
}) {
final fixed = _fixed;
if (fixed == null) {
_fixed = sampling;
_fixedExplicit = explicit;
return null;
}
if (fixed == sampling) return null;
final message =
'this session asks for $sampling, but LiteRT-LM keeps the sampler of '
"the engine's first generation, $fixed, for every later conversation "
'(google-ai-edge/LiteRT-LM#2080). Close and reload the model to change '
'it.';
if (!explicit && !_fixedExplicit) return (message: message, warn: false);
if (_warned) return null;
_warned = true;
return (message: message, warn: true);
}