mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-25 02:44:36 +00:00
server : do not speculate during prompt processing
ggml-ci
This commit is contained in:
parent
642330ac7c
commit
33d7b70c88
@ -2322,6 +2322,10 @@ struct server_context {
|
||||
continue;
|
||||
}
|
||||
|
||||
if (slot.state != SLOT_STATE_GENERATING) {
|
||||
continue;
|
||||
}
|
||||
|
||||
llama_token id = slot.sampled;
|
||||
|
||||
struct common_speculative_params params_spec;
|
||||
|
Loading…
Reference in New Issue
Block a user