release: 1.78.0

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "1.77.0"
+  ".": "1.78.0"
 }

.stats.yml

@@ -1,4 +1,4 @@
-configured_endpoints: 97
-openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-0ee6b36cf3cc278cef4199a6aec5f7d530a6c1f17a74830037e96d50ca1edc50.yml
-openapi_spec_hash: e8ec5f46bc0655b34f292422d58a60f6
-config_hash: d9b6b6e6bc85744663e300eebc482067
+configured_endpoints: 101
+openapi_spec_url: https://storage.googleapis.com/stainless-sdk-openapi-specs/openai%2Fopenai-794a6ed3c3d3d77887564755168056af8a426b17cf1ec721e3a300503dc22a41.yml
+openapi_spec_hash: 25a81c220713cd5b0bafc221d1dfa79a
+config_hash: 0b768ed1b56c6d82816f0fa40dc4aaf5

CHANGELOG.md

@@ -1,5 +1,25 @@
 # Changelog
 
+## 1.78.0 (2025-05-08)
+
+Full Changelog: [v1.77.0...v1.78.0](https://github.com/openai/openai-python/compare/v1.77.0...v1.78.0)
+
+### Features
+
+* **api:** Add reinforcement fine-tuning api support ([bebe361](https://github.com/openai/openai-python/commit/bebe36104bd3062d09ab9bbfb4bacfc99e737cb2))
+
+
+### Bug Fixes
+
+* ignore errors in isinstance() calls on LazyProxy subclasses ([#2343](https://github.com/openai/openai-python/issues/2343)) ([52cbbdf](https://github.com/openai/openai-python/commit/52cbbdf2207567741f16d18f1ea1b0d13d667375)), closes [#2056](https://github.com/openai/openai-python/issues/2056)
+
+
+### Chores
+
+* **internal:** update proxy tests ([b8e848d](https://github.com/openai/openai-python/commit/b8e848d5fb58472cbfa27fb3ed01efc25a05d944))
+* use lazy imports for module level client ([4d0f409](https://github.com/openai/openai-python/commit/4d0f409e79a18cce9855fe076f5a50e52b8bafd8))
+* use lazy imports for resources ([834813c](https://github.com/openai/openai-python/commit/834813c5cb1a84effc34e5eabed760393e1de806))
+
 ## 1.77.0 (2025-05-02)
 
 Full Changelog: [v1.76.2...v1.77.0](https://github.com/openai/openai-python/compare/v1.76.2...v1.77.0)

api.md

@@ -225,6 +225,21 @@ Methods:
 
 # FineTuning
 
+## Methods
+
+Types:
+
+```python
+from openai.types.fine_tuning import (
+    DpoHyperparameters,
+    DpoMethod,
+    ReinforcementHyperparameters,
+    ReinforcementMethod,
+    SupervisedHyperparameters,
+    SupervisedMethod,
+)
+```
+
 ## Jobs
 
 Types:
@@ -246,6 +261,8 @@ Methods:
 - <code title="get /fine_tuning/jobs">client.fine_tuning.jobs.<a href="./src/openai/resources/fine_tuning/jobs/jobs.py">list</a>(\*\*<a href="src/openai/types/fine_tuning/job_list_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/fine_tuning_job.py">SyncCursorPage[FineTuningJob]</a></code>
 - <code title="post /fine_tuning/jobs/{fine_tuning_job_id}/cancel">client.fine_tuning.jobs.<a href="./src/openai/resources/fine_tuning/jobs/jobs.py">cancel</a>(fine_tuning_job_id) -> <a href="./src/openai/types/fine_tuning/fine_tuning_job.py">FineTuningJob</a></code>
 - <code title="get /fine_tuning/jobs/{fine_tuning_job_id}/events">client.fine_tuning.jobs.<a href="./src/openai/resources/fine_tuning/jobs/jobs.py">list_events</a>(fine_tuning_job_id, \*\*<a href="src/openai/types/fine_tuning/job_list_events_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/fine_tuning_job_event.py">SyncCursorPage[FineTuningJobEvent]</a></code>
+- <code title="post /fine_tuning/jobs/{fine_tuning_job_id}/pause">client.fine_tuning.jobs.<a href="./src/openai/resources/fine_tuning/jobs/jobs.py">pause</a>(fine_tuning_job_id) -> <a href="./src/openai/types/fine_tuning/fine_tuning_job.py">FineTuningJob</a></code>
+- <code title="post /fine_tuning/jobs/{fine_tuning_job_id}/resume">client.fine_tuning.jobs.<a href="./src/openai/resources/fine_tuning/jobs/jobs.py">resume</a>(fine_tuning_job_id) -> <a href="./src/openai/types/fine_tuning/fine_tuning_job.py">FineTuningJob</a></code>
 
 ### Checkpoints
 
@@ -279,6 +296,38 @@ Methods:
 - <code title="get /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">retrieve</a>(fine_tuned_model_checkpoint, \*\*<a href="src/openai/types/fine_tuning/checkpoints/permission_retrieve_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_retrieve_response.py">PermissionRetrieveResponse</a></code>
 - <code title="delete /fine_tuning/checkpoints/{fine_tuned_model_checkpoint}/permissions/{permission_id}">client.fine_tuning.checkpoints.permissions.<a href="./src/openai/resources/fine_tuning/checkpoints/permissions.py">delete</a>(permission_id, \*, fine_tuned_model_checkpoint) -> <a href="./src/openai/types/fine_tuning/checkpoints/permission_delete_response.py">PermissionDeleteResponse</a></code>
 
+## Alpha
+
+### Graders
+
+Types:
+
+```python
+from openai.types.fine_tuning.alpha import GraderRunResponse, GraderValidateResponse
+```
+
+Methods:
+
+- <code title="post /fine_tuning/alpha/graders/run">client.fine_tuning.alpha.graders.<a href="./src/openai/resources/fine_tuning/alpha/graders.py">run</a>(\*\*<a href="src/openai/types/fine_tuning/alpha/grader_run_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/alpha/grader_run_response.py">GraderRunResponse</a></code>
+- <code title="post /fine_tuning/alpha/graders/validate">client.fine_tuning.alpha.graders.<a href="./src/openai/resources/fine_tuning/alpha/graders.py">validate</a>(\*\*<a href="src/openai/types/fine_tuning/alpha/grader_validate_params.py">params</a>) -> <a href="./src/openai/types/fine_tuning/alpha/grader_validate_response.py">GraderValidateResponse</a></code>
+
+# Graders
+
+## GraderModels
+
+Types:
+
+```python
+from openai.types.graders import (
+    LabelModelGrader,
+    MultiGrader,
+    PythonGrader,
+    ScoreModelGrader,
+    StringCheckGrader,
+    TextSimilarityGrader,
+)
+```
+
 # VectorStores
 
 Types:
@@ -738,10 +787,7 @@ Types:
 ```python
 from openai.types import (
     EvalCustomDataSourceConfig,
-    EvalLabelModelGrader,
     EvalStoredCompletionsDataSourceConfig,
-    EvalStringCheckGrader,
-    EvalTextSimilarityGrader,
     EvalCreateResponse,
     EvalRetrieveResponse,
     EvalUpdateResponse,

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "openai"
-version = "1.77.0"
+version = "1.78.0"
 description = "The official Python library for the openai API"
 dynamic = ["readme"]
 license = "Apache-2.0"