mirror of
https://github.com/dogkeeper886/ollama37.git
synced 2025-12-10 07:46:59 +00:00
llm: add back check for empty token cache
This commit is contained in:
2
llm/ext_server/server.cpp
vendored
2
llm/ext_server/server.cpp
vendored
@@ -1032,7 +1032,7 @@ struct llama_server_context
|
|||||||
slot.has_next_token = false;
|
slot.has_next_token = false;
|
||||||
}
|
}
|
||||||
|
|
||||||
if (llama_token_is_eog(model, result.tok))
|
if (!slot.cache_tokens.empty() && llama_token_is_eog(model, result.tok))
|
||||||
{
|
{
|
||||||
slot.stopped_eos = true;
|
slot.stopped_eos = true;
|
||||||
slot.has_next_token = false;
|
slot.has_next_token = false;
|
||||||
|
|||||||
Reference in New Issue
Block a user