fix(agentcore): remove lazy loading to resolve circular import deadlock

The lazy loading system introduced in commit 6b6fbad created a circular
import deadlock that prevented the FastAPI proxy server from starting:

1. proxy_server.py:56 imports from litellm.utils
2. utils.py:56 imports litellm module
3. litellm/__init__.py sets up __getattr__ for lazy loading
4. Later code accesses litellm.ModelResponse
5. __getattr__ tries to import from utils again
6. DEADLOCK: utils.py is still being imported from step 2

This caused ECS health checks to fail because Uvicorn never started.

Solution:
- Restored direct imports for all utils functions (30+ functions)
- Restored direct imports for cost_calculator functions
- Restored direct imports for litellm_logging functions
- Removed __getattr__ lazy loading system
- Removed TYPE_CHECKING stub declarations

This reverts the lazy loading optimization while maintaining all
functionality. The module now imports successfully and the proxy
server can start without hanging.

Tested:
✅ litellm module imports without hanging
✅ All functions accessible (ModelResponse, completion_cost, Logging)
✅ proxy_server.py imports successfully (no circular import)

Author: busla

litellm/__init__.py

@@ -444,6 +444,8 @@
 output_parse_pii: bool = False
 #############################################
 from litellm.litellm_core_utils.get_model_cost_map import get_model_cost_map
+from .cost_calculator import completion_cost, cost_per_token, response_cost_calculator
+from litellm.litellm_core_utils.litellm_logging import Logging, modify_integration
 
 model_cost = get_model_cost_map(url=model_cost_map_url)
 cost_discount_config: Dict[str, float] = (
@@ -1058,9 +1060,49 @@ def add_known_models():
 from litellm.litellm_core_utils.core_helpers import remove_index_from_tool_calls
 from litellm.litellm_core_utils.token_counter import get_modified_max_tokens
 # client must be imported immediately as it's used as a decorator at function definition time
-from .utils import client
-# Note: Most other utils imports are lazy-loaded via __getattr__ to avoid loading utils.py 
-# (which imports tiktoken) at import time
+from .utils import (
+    client,
+    exception_type,
+    get_optional_params,
+    get_response_string,
+    token_counter,
+    create_pretrained_tokenizer,
+    create_tokenizer,
+    supports_function_calling,
+    supports_web_search,
+    supports_url_context,
+    supports_response_schema,
+    supports_parallel_function_calling,
+    supports_vision,
+    supports_audio_input,
+    supports_audio_output,
+    supports_system_messages,
+    supports_reasoning,
+    get_litellm_params,
+    acreate,
+    get_max_tokens,
+    get_model_info,
+    register_prompt_template,
+    validate_environment,
+    check_valid_key,
+    register_model,
+    encode,
+    decode,
+    _calculate_retry_after,
+    _should_retry,
+    get_supported_openai_params,
+    get_api_base,
+    get_first_chars_messages,
+    ModelResponse,
+    ModelResponseStream,
+    EmbeddingResponse,
+    ImageResponse,
+    TranscriptionResponse,
+    TextCompletionResponse,
+    get_provider_fields,
+    ModelResponseListIterator,
+    get_valid_models,
+)
 
 from .llms.bytez.chat.transformation import BytezChatConfig
 from .llms.custom_llm import CustomLLM
@@ -1491,56 +1533,3 @@ def set_global_gitlab_config(config: Dict[str, Any]) -> None:
     global_gitlab_config = config
 
 
-# Lazy loading system for heavy modules to reduce initial import time and memory usage
-
-if TYPE_CHECKING:
-    cost_per_token: Callable[..., Tuple[float, float]]
-    completion_cost: Callable[..., float]
-    response_cost_calculator: Any
-    modify_integration: Any
-
-
-def __getattr__(name: str) -> Any:
-    """Lazy import handler for cost_calculator and litellm_logging functions."""
-    # Lazy load cost_calculator functions
-    _cost_calculator_names = (
-        "completion_cost",
-        "cost_per_token",
-        "response_cost_calculator",
-    )
-    if name in _cost_calculator_names:
-        from ._lazy_imports import _lazy_import_cost_calculator
-        return _lazy_import_cost_calculator(name)
-    
-    # Lazy load litellm_logging functions
-    _litellm_logging_names = (
-        "Logging",
-        "modify_integration",
-    )
-    if name in _litellm_logging_names:
-        from ._lazy_imports import _lazy_import_litellm_logging
-        return _lazy_import_litellm_logging(name)
-    
-    # Lazy load utils functions
-    _utils_names = (
-        "exception_type", "get_optional_params", "get_response_string", "token_counter",
-        "create_pretrained_tokenizer", "create_tokenizer", "supports_function_calling",
-        "supports_web_search", "supports_url_context", "supports_response_schema",
-        "supports_parallel_function_calling", "supports_vision", "supports_audio_input",
-        "supports_audio_output", "supports_system_messages", "supports_reasoning",
-        "get_litellm_params", "acreate", "get_max_tokens", "get_model_info",
-        "register_prompt_template", "validate_environment", "check_valid_key",
-        "register_model", "encode", "decode", "_calculate_retry_after", "_should_retry",
-        "get_supported_openai_params", "get_api_base", "get_first_chars_messages",
-        "ModelResponse", "ModelResponseStream", "EmbeddingResponse", "ImageResponse",
-        "TranscriptionResponse", "TextCompletionResponse", "get_provider_fields",
-        "ModelResponseListIterator", "get_valid_models",
-    )
-    if name in _utils_names:
-        from ._lazy_imports import _lazy_import_utils
-        return _lazy_import_utils(name)
-    
-    raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
-
-
-# ALL_LITELLM_RESPONSE_TYPES is lazy-loaded via __getattr__ to avoid loading utils at import time