fix: Update default chat model to grok-4-1-fast-reasoning and enhance logging for LLM responses

This commit is contained in:
bsiggel
2026-03-19 09:50:31 +00:00
parent 7fffdb2660
commit 2ac83df1e0
2 changed files with 15 additions and 3 deletions

View File

@@ -17,7 +17,7 @@ class LangChainXAIService:
Usage: Usage:
service = LangChainXAIService(ctx) service = LangChainXAIService(ctx)
model = service.get_chat_model(model="grok-2-latest") model = service.get_chat_model(model="grok-4-1-fast-reasoning")
model_with_tools = service.bind_file_search(model, collection_id) model_with_tools = service.bind_file_search(model, collection_id)
result = await service.invoke_chat(model_with_tools, messages) result = await service.invoke_chat(model_with_tools, messages)
""" """
@@ -46,7 +46,7 @@ class LangChainXAIService:
def get_chat_model( def get_chat_model(
self, self,
model: str = "grok-2-latest", model: str = "grok-4-1-fast-reasoning",
temperature: float = 0.7, temperature: float = 0.7,
max_tokens: Optional[int] = None max_tokens: Optional[int] = None
): ):
@@ -54,7 +54,7 @@ class LangChainXAIService:
Initialisiert ChatXAI Model. Initialisiert ChatXAI Model.
Args: Args:
model: Model name (default: grok-2-latest) model: Model name (default: grok-4-1-fast-reasoning)
temperature: Sampling temperature 0.0-1.0 temperature: Sampling temperature 0.0-1.0
max_tokens: Optional max tokens for response max_tokens: Optional max tokens for response

View File

@@ -317,6 +317,18 @@ async def invoke_and_format_response(
"total_tokens": getattr(u, 'input_tokens', 0) + getattr(u, 'output_tokens', 0) "total_tokens": getattr(u, 'input_tokens', 0) + getattr(u, 'output_tokens', 0)
} }
# Log complete LLM response
ctx.logger.info("=" * 80)
ctx.logger.info("📤 LLM RESPONSE")
ctx.logger.info("-" * 80)
ctx.logger.info(f"Model: {model_name}")
ctx.logger.info(f"Completion ID: {completion_id}")
ctx.logger.info(f"Usage: {usage['prompt_tokens']} prompt + {usage['completion_tokens']} completion = {usage['total_tokens']} total tokens")
ctx.logger.info("-" * 80)
ctx.logger.info("Content:")
ctx.logger.info(content)
ctx.logger.info("=" * 80)
# Format OpenAI-compatible response # Format OpenAI-compatible response
response_body = { response_body = {
'id': completion_id, 'id': completion_id,