frequency penalty fixes

src/leapfrogai/cli.py

@@ -1,8 +1,10 @@
 import click
 import sys
+import asyncio
 
 from leapfrogai.errors import AppImportError
 from leapfrogai.utils import import_app
+from leapfrogai.serve import serve
 
 @click.argument("app", envvar="LEAPFROGAI_APP")
 @click.option(
@@ -31,7 +33,7 @@ def cli(app: str, host: str, port: str, app_dir: str):
     sys.path.insert(0, app_dir)
     """Leapfrog AI CLI"""
     app = import_app(app)
-    app().serve(host, port)
+    asyncio.run(serve(app(), host, port))
 
 
 if __name__ == "__main__":

src/leapfrogai/llm.py

@@ -31,7 +31,7 @@ class GenerationConfig(BaseModel):
     stop: List[str]
     repetition_penalty: float
     presence_penalty: float
-    frequency_penalty: float
+    frequency_penalty: float | None = None
     best_of: str
     logit_bias: dict[str, int]
     return_full_text: bool
@@ -58,10 +58,9 @@ def _build_gen_stream(self, prompt: str, request: ChatCompletionRequest | Comple
                 top_p=request.top_p,
                 do_sample=request.do_sample,
                 n=request.n,
-                stop=request.stop,
+                stop=list(request.stop),
                 repetition_penalty=request.repetition_penalty,
                 presence_penalty=request.presence_penalty,
-                frequency_penalty=request.frequency_penalty,
                 best_of=request.best_of,
                 logit_bias=request.logit_bias,
                 return_full_text=request.return_full_text,
@@ -107,9 +106,6 @@ async def CompleteStream(self, request: CompletionRequest, context: GrpcContext)
             for text_chunk in gen_stream:
                 choice = CompletionChoice(index=0, text=text_chunk)
                 yield CompletionResponse(choices=[choice])
-
-        def serve(self, host, port):
-            asyncio.run(serve(self, host, port))
 
     NewClass.__name__ = _cls.__name__
     return NewClass