server : (tests) don't use thread for capturing stdout/stderr, bump openai client library (ggml-org#10568)

* server : (tests) don't use thread for capturing stdout/stderr

* test: bump openai to 1.55.2

* bump openai to 1.55.3

Author: ngxson

examples/server/tests/requirements.txt

@@ -2,6 +2,6 @@ aiohttp~=3.9.3
 pytest~=8.3.3
 huggingface_hub~=0.23.2
 numpy~=1.26.4
-openai~=1.30.3
+openai~=1.55.3
 prometheus-client~=0.20.0
 requests~=2.32.3

examples/server/tests/utils.py

@@ -8,7 +8,6 @@
 import re
 import json
 import sys
-import threading
 import requests
 import time
 from concurrent.futures import ThreadPoolExecutor, as_completed
@@ -161,26 +160,12 @@ def start(self, timeout_seconds: int = 10) -> None:
         self.process = subprocess.Popen(
             [str(arg) for arg in [server_path, *server_args]],
             creationflags=flags,
-            stdout=subprocess.PIPE,
-            stderr=subprocess.PIPE,
+            stdout=sys.stdout,
+            stderr=sys.stdout,
             env={**os.environ, "LLAMA_CACHE": "tmp"},
         )
         server_instances.add(self)
 
-        def server_log(in_stream, out_stream):
-            for line in iter(in_stream.readline, b""):
-                print(line.decode("utf-8"), end="", file=out_stream)
-
-        thread_stdout = threading.Thread(
-            target=server_log, args=(self.process.stdout, sys.stdout), daemon=True
-        )
-        thread_stdout.start()
-
-        thread_stderr = threading.Thread(
-            target=server_log, args=(self.process.stderr, sys.stderr), daemon=True
-        )
-        thread_stderr.start()
-
         print(f"server pid={self.process.pid}, pytest pid={os.getpid()}")
 
         # wait for server to start