dearwolf/aider
1
0
Fork 0
mirror of https://github.com/Aider-AI/aider.git synced 2025-05-28 16:25:00 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

refactor: Defer litellm registration to improve loading performance

Browse source
This commit is contained in:
Paul Gauthier (aider) 2025-02-04 15:51:55 -08:00 • committed by Paul Gauthier
parent 44365651a6
commit 5ad8bb1830
1 changed files with 10 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

13
aider/models.py
View file

@ -85,6 +85,8 @@ MODEL_ALIASES = {
"r1": "deepseek/deepseek-reasoner",
"flash": "gemini/gemini-2.0-flash-exp",
}
# Deferred model definitions loaded from metadata files
LITELLM_MODEL_DEFS = []
@dataclass
@ -238,6 +240,12 @@ class Model(ModelSettings):
self.get_editor_model(editor_model, editor_edit_format)
def get_model_info(self, model):
global LITELLM_MODEL_DEFS
if LITELLM_MODEL_DEFS:
litellm._load_litellm()
for model_def in LITELLM_MODEL_DEFS:
litellm.register_model(model_def)
LITELLM_MODEL_DEFS.clear()
return model_info_manager.get_model_info(model)
def _copy_fields(self, source):
@ -665,9 +673,8 @@ def register_litellm_models(model_fnames):
if not model_def:
continue
# only load litellm if we have actual data
litellm._load_litellm()
litellm.register_model(model_def)
# Defer registration with litellm to faster path.
LITELLM_MODEL_DEFS.append(model_def)
except Exception as e:
raise Exception(f"Error loading model definition from {model_fname}: {e}")