fix(minor): Minor ruff fixes

Author: abetlen

llama_cpp/_internals.py

@@ -4,6 +4,7 @@
 import ctypes
 
 from typing import (
+    Dict,
     List,
     Optional,
     Sequence,

llama_cpp/llama.py

@@ -11,10 +11,11 @@
 import warnings
 import contextlib
 import multiprocessing
-from types import TracebackType
 
 from typing import (
+    Any,
     List,
+    Literal,
     Optional,
     Union,
     Generator,
@@ -23,14 +24,11 @@
     Deque,
     Callable,
     Dict,
-    Type,
 )
 from collections import deque
 from pathlib import Path
 
 
-from llama_cpp.llama_types import List
-
 from .llama_types import *
 from .llama_grammar import LlamaGrammar
 from .llama_cache import (
@@ -901,7 +899,7 @@ def embed(
         pooling_type = self.pooling_type()
         logits_all = pooling_type == llama_cpp.LLAMA_POOLING_TYPE_NONE
 
-        if self.context_params.embeddings == False:
+        if self.context_params.embeddings is False:
             raise RuntimeError(
                 "Llama model must be created with embedding=True to call this method"
             )

llama_cpp/llama_chat_format.py

@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import os
+import sys
 import json
 import ctypes
 import dataclasses
@@ -627,6 +628,8 @@ def chat_completion_handler(
                     json.dumps(schema), verbose=llama.verbose
                 )
             except Exception as e:
+                if llama.verbose:
+                    print(str(e), file=sys.stderr)
                 grammar = llama_grammar.LlamaGrammar.from_string(
                     llama_grammar.JSON_GBNF, verbose=llama.verbose
                 )
@@ -1611,12 +1614,12 @@ def message_to_str(msg: llama_types.ChatCompletionRequestMessage):
         function_call = completion_text.split(".")[-1][:-1]
         new_prompt = prompt + completion_text + stop
     elif isinstance(function_call, str) and function_call != "none":
-        new_prompt = prompt + f":\n"
+        new_prompt = prompt + ":\n"
     elif isinstance(function_call, dict):
         new_prompt = prompt + f" to=functions.{function_call['name']}:\n"
         function_call = function_call["name"]
     else:
-        new_prompt = prompt + f":\n"
+        new_prompt = prompt + ":\n"
 
     function_body = None
     for function in functions or []:
@@ -2871,6 +2874,8 @@ def embed_image_bytes(image_bytes: bytes):
                     json.dumps(schema), verbose=llama.verbose
                 )
             except Exception as e:
+                if llama.verbose:
+                    print(str(e), file=sys.stderr)
                 grammar = llama_grammar.LlamaGrammar.from_string(
                     llama_grammar.JSON_GBNF, verbose=llama.verbose
                 )