Modulesยง
- agent_
applicable_ state - agent_
applicable_ state_ holder - agent_
desired_ state_ converter - agent_
issue_ fix - agent_
kv_ cache_ dtype - agent_
pooling_ type - chat_
template_ renderer - continue_
from_ conversation_ history_ request - continue_
from_ raw_ prompt_ request - continuous_
batch_ active_ request - continuous_
batch_ arbiter - continuous_
batch_ arbiter_ build_ outcome - continuous_
batch_ arbiter_ handle - continuous_
batch_ embedding_ processor - continuous_
batch_ request_ phase - continuous_
batch_ request_ state - continuous_
batch_ scheduler - continuous_
batch_ scheduler_ command - continuous_
batch_ scheduler_ context - converts_
to_ llama_ kv_ cache_ dtype - converts_
to_ llama_ pooling_ type - decoded_
image - decoded_
image_ error - desired_
model_ resolution - dispenses_
slots - drain_
in_ flight_ requests - embedding_
input_ tokenized - generate_
embedding_ batch_ request - grammar_
sampler - llamacpp_
arbiter_ service - management_
socket_ client_ service - model_
metadata_ holder - model_
source - normalization
- plan_
embedding_ batches - prepare_
conversation_ history_ request - prepared_
conversation_ history_ request - receive_
stream_ stopper_ collection - reconciliation_
service - resolve_
desired_ model - resolve_
grammar - resolve_
grammar_ to_ gbnf - resolved_
grammar - resolves_
model_ source - sample_
token_ at_ batch_ index - sampling_
outcome - sequence_
id_ pool - slot_
aggregated_ status - slot_
aggregated_ status_ download_ progress - slot_
aggregated_ status_ manager - slot_
guard - tool_
call_ buffer - tool_
call_ event - tool_
call_ pipeline - tool_
call_ pipeline_ error - tool_
call_ validator - validator_
build_ error