Add /v1/internal/model/load endpoint (tentative)

extensions/openai/models.py

@@ -1,4 +1,6 @@
 from modules import shared
+from modules.models import load_model, unload_model
+from modules.models_settings import get_model_metadata, update_model_parameters
 from modules.utils import get_available_models
 
 
@@ -35,3 +37,27 @@ def get_dummy_models() -> list:
         'gpt-3.5-turbo',
         'text-embedding-ada-002',
     ]
+
+
+def _load_model(data):
+    model_name = data["model_name"]
+    args = data["args"]
+    settings = data["settings"]
+
+    unload_model()
+    model_settings = get_model_metadata(model_name)
+    update_model_parameters(model_settings, initial=True)
+
+    # Update shared.args with custom model loading settings
+    if args:
+        for k in args:
+            if k in shared.args:
+                setattr(shared.args, k, args[k])
+
+    shared.model, shared.tokenizer = load_model(model_name)
+
+    # Update shared.settings with custom generation defaults
+    if settings:
+        for k in settings:
+            if k in shared.settings:
+                shared.settings[k] = settings[k]

extensions/openai/script.py

@@ -1,5 +1,6 @@
 import json
 import os
+import traceback
 from threading import Thread
 
 import extensions.openai.completions as OAIcompletions
@@ -31,6 +32,7 @@ from .typing import (
     DecodeResponse,
     EncodeRequest,
     EncodeResponse,
+    LoadModelRequest,
     ModelInfoResponse,
     TokenCountResponse,
     to_dict
@@ -231,12 +233,22 @@ async def handle_stop_generation(request: Request):
     return JSONResponse(content="OK")
 
 
-@app.get("/v1/internal/model-info", response_model=ModelInfoResponse)
+@app.get("/v1/internal/model/info", response_model=ModelInfoResponse)
 async def handle_model_info():
     payload = OAImodels.get_current_model_info()
     return JSONResponse(content=payload)
 
 
+@app.post("/v1/internal/model/load")
+async def handle_load_model(request_data: LoadModelRequest):
+    try:
+        OAImodels._load_model(to_dict(request_data))
+        return JSONResponse(content="OK")
+    except:
+        traceback.print_exc()
+        return HTTPException(status_code=400, detail="Failed to load the model.")
+
+
 def run_server():
     server_addr = '0.0.0.0' if shared.args.listen else '127.0.0.1'
     port = int(os.environ.get('OPENEDAI_PORT', shared.args.api_port))

extensions/openai/typing.py

@@ -147,6 +147,12 @@ class ModelInfoResponse(BaseModel):
     lora_names: List[str]
 
 
+class LoadModelRequest(BaseModel):
+    model_name: str
+    args: dict | None = None
+    settings: dict | None = None
+
+
 def to_json(obj):
     return json.dumps(obj.__dict__, indent=4)
 

modules/models.py

@@ -79,7 +79,7 @@ def load_model(model_name, loader=None):
             loader = metadata['loader']
             if loader is None:
                 logger.error('The path to the model does not exist. Exiting.')
-                return None, None
+                raise ValueError
 
     shared.args.loader = loader
     output = load_func_map[loader](model_name)

server.py

@@ -216,8 +216,7 @@ if __name__ == "__main__":
             model_name = shared.model_name
 
         model_settings = get_model_metadata(model_name)
-        shared.settings.update({k: v for k, v in model_settings.items() if k in shared.settings})  # hijacking the interface defaults
-        update_model_parameters(model_settings, initial=True)  # hijacking the command-line arguments
+        update_model_parameters(model_settings, initial=True)  # hijack the command-line arguments
 
         # Load the model
         shared.model, shared.tokenizer = load_model(model_name)