release: 1.2.4

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "1.2.3"
+  ".": "1.2.4"
 }

CHANGELOG.md

@@ -1,5 +1,13 @@
 # Changelog
 
+## 1.2.4 (2023-11-13)
+
+Full Changelog: [v1.2.3...v1.2.4](https://github.com/openai/openai-python/compare/v1.2.3...v1.2.4)
+
+### Bug Fixes
+
+* **client:** retry if SSLWantReadError occurs in the async client ([#804](https://github.com/openai/openai-python/issues/804)) ([be82288](https://github.com/openai/openai-python/commit/be82288f3c88c10c9ac20ba3b8cb53b5c7a4e2f9))
+
 ## 1.2.3 (2023-11-10)
 
 Full Changelog: [v1.2.2...v1.2.3](https://github.com/openai/openai-python/compare/v1.2.2...v1.2.3)

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "openai"
-version = "1.2.3"
+version = "1.2.4"
 description = "The official Python library for the openai API"
 readme = "README.md"
 license = "Apache-2.0"

src/openai/_base_client.py

@@ -1320,12 +1320,6 @@ async def _request(
             if retries > 0:
                 return await self._retry_request(options, cast_to, retries, stream=stream, stream_cls=stream_cls)
             raise APITimeoutError(request=request) from err
-        except httpx.ReadTimeout as err:
-            # We explicitly do not retry on ReadTimeout errors as this means
-            # that the server processing the request has taken 60 seconds
-            # (our default timeout). This likely indicates that something
-            # is not working as expected on the server side.
-            raise
         except httpx.TimeoutException as err:
             if retries > 0:
                 return await self._retry_request(options, cast_to, retries, stream=stream, stream_cls=stream_cls)

src/openai/_version.py

@@ -1,4 +1,4 @@
 # File generated from our OpenAPI spec by Stainless.
 
 __title__ = "openai"
-__version__ = "1.2.3"  # x-release-please-version
+__version__ = "1.2.4"  # x-release-please-version