feat(*): support ark batch chat

Author: jizhongBao

volcenginesdkarkruntime/_base_client.py

@@ -843,7 +843,6 @@ async def delete(
 
         return await self.request(cast_to, opts)
 
-
     async def post_without_retry(
             self,
             path: str,

volcenginesdkarkruntime/resources/batch_chat/completions.py

@@ -1,12 +1,11 @@
 from __future__ import annotations
 
 import time
-from datetime import timedelta,datetime
+from datetime import timedelta, datetime
 from random import random
-from typing import Dict, List, Union, Iterable, Optional, Callable, Iterator, AsyncIterator
+from typing import Dict, List, Union, Iterable, Optional, Callable
 
 import httpx
-from typing_extensions import Literal
 
 from ..._exceptions import ArkAPITimeoutError, ArkAPIConnectionError, ArkAPIStatusError
 from ..._types import Body, Query, Headers
@@ -30,14 +29,14 @@
 from ..._constants import (
     ARK_E2E_ENCRYPTION_HEADER,
     INITIAL_RETRY_DELAY,
-    MAX_RETRY_DELAY, DEFAULT_TIMEOUT
+    MAX_RETRY_DELAY,
 )
 
 __all__ = ["Completions", "AsyncCompletions"]
 
 
 def _calculate_retry_timeout(retry_times) -> float:
-    nbRetries = min(retry_times, MAX_RETRY_DELAY/INITIAL_RETRY_DELAY)
+    nbRetries = min(retry_times, MAX_RETRY_DELAY / INITIAL_RETRY_DELAY)
     sleep_seconds = min(INITIAL_RETRY_DELAY * pow(2, nbRetries), MAX_RETRY_DELAY)
     # Apply some jitter, plus-or-minus half a second.
     jitter = 1 - 0.25 * random()
@@ -102,7 +101,6 @@ def _encrypt(self, model: str, messages: Iterable[ChatCompletionMessageParam], e
                                                                                   x))
         return _crypto_key, _crypto_nonce
 
-
     def _decrypt(self, key: bytes, nonce: bytes, resp: ChatCompletion
                  ) -> ChatCompletion:
         if resp.choices is not None:
@@ -241,7 +239,6 @@ async def _decrypt(self, key: bytes, nonce: bytes, resp: ChatCompletion
                 resp.choices[index] = choice
         return resp
 
-
     @async_with_sts_token
     async def create(
         self,

volcenginesdkexamples/volcenginesdkarkruntime/async_batch_chat_completions.py

@@ -15,6 +15,7 @@
 # For more information，please check this document（https://www.volcengine.com/docs/82379/1263279）
 client = AsyncArk()
 
+
 async def worker(semaphore, worker_id, task_num):
     async with semaphore:
         print(f"Worker {worker_id} is starting.")
@@ -35,6 +36,7 @@ async def worker(semaphore, worker_id, task_num):
                 print(f"Worker {worker_id} task {i} is completed.")
         print(f"Worker {worker_id} is completed.")
 
+
 async def main():
     start = datetime.now()
     max_concurrent_tasks = 1000