[Core] Add max_num_queued_reqs and max_num_queued_tokens for queue size management (#49445)
Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
This commit is contained in:
@@ -56,8 +56,7 @@ class OpenAIServingChatBatch(OpenAIServingChat):
|
||||
logger.error("Error with model %s", error_check_ret)
|
||||
return error_check_ret
|
||||
|
||||
if self.engine_client.errored:
|
||||
raise self.engine_client.dead_error
|
||||
self._preflight()
|
||||
|
||||
renderer = self.online_renderer
|
||||
|
||||
|
||||
Reference in New Issue
Block a user