WARNING: All log messages before absl::InitializeLog() is called are written to STDERR

[LindaLawton]

Description of the bug:

I updated to the latest version of the pip package.

from dotenv import load_dotenv
import os
import google.generativeai as genai
from helpers import constants

load_dotenv()

# name of the AI model used in this call.
TEXT_MODEL_NAME_LATEST = os.getenv("TEXT_MODEL_NAME_LATEST")
PDF_PAGE_IMAGES_PATH = os.getenv("PDF_PAGE_IMAGES_PATH")
PROMPT_PATH = constants.PROMPT_PATH

class GeminiService:
    generation_config_json: str = {
        'temperature': 0.9,
        'top_p': 1,
        'top_k': 40,
        'max_output_tokens': 2048,
        'stop_sequences': [],
        "response_mime_type": "application/json",
    }

    generation_config: str = {
        'temperature': 0.9,
        'top_p': 1,
        'top_k': 40,
        'max_output_tokens': 2048,
        'stop_sequences': [],
    }

    safety_settings: list[str] = [{"category": "HARM_CATEGORY_HARASSMENT", "threshold": "BLOCK_NONE"},
                                  {"category": "HARM_CATEGORY_HATE_SPEECH", "threshold": "BLOCK_NONE"},
                                  {"category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "threshold": "BLOCK_NONE"},
                                  {"category": "HARM_CATEGORY_DANGEROUS_CONTENT", "threshold": "BLOCK_NONE"}]

    def __init__(self, prompt_path, as_json: bool = True):
        genai.configure(api_key=os.getenv("API_KEY"))
        self.model = genai.GenerativeModel(model_name=os.getenv("TEXT_MODEL_NAME_LATEST"),
                                           system_instruction=GeminiService.read_prompt(prompt_path),
                                           generation_config=GeminiService.generation_config_json if as_json else GeminiService.generation_config,
                                           safety_settings=self.safety_settings)

    @staticmethod
    def read_prompt(prompt_file):
        with open(prompt_file, 'r', encoding="utf-8") as file:
            # Load the JSON data from the file into a Python dictionary
            return file.read()

    def single_completion(self, request) -> str:
        return self.model.generate_content(request).text

if __name__ == '__main__':
    service = GeminiService( os.path.join("..", constants.PROMPT_PATH, "prompt.txt"))
    prompt = "What is the meaning of life"
    response = service.single_completion(prompt)
    print(response)

My code is now giving a strange warning.

WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
I0000 00:00:1721829902.342524    9100 config.cc:230] gRPC experiments enabled: call_status_override_on_cancellation, event_engine_client, event_engine_dns, event_engine_listener, http2_stats_fix, monitoring_experiment, pick_first_new, trace_record_callops, work_serializer_clears_time_cache

It did not do this before nor does it in another project

Actual vs expected behavior:

not give weird warning

Any other information you'd like to share?

What do i need to do to get ride of this?

[MarkDaoust]

Hi, It looks like this is an issue with GRPC and protobuf: https://github.com/grpc/grpc/issues/37222

If I upgrade to pip install -U grpcio I get this warning, if I use pip install grpcio==1.60.1 it's gone.

This is not something I can fix here. You can wait for the fixed version of grpcio, or wait for the fixed version.

[MarkDaoust]

PDF_PAGE_IMAGES_PATH

Note that we did just add PDF support: https://github.com/google-gemini/cookbook/blob/main/quickstarts/PDF_Files.ipynb