verifiers 0.2.1.dev3__tar.gz → 0.2.1.dev4__tar.gz

{verifiers-0.2.1.dev3 → verifiers-0.2.1.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: verifiers
-Version: 0.2.1.dev3
+Version: 0.2.1.dev4
 Summary: Verifiers: Environments for LLM Reinforcement Learning
 Project-URL: Homepage, https://github.com/primeintellect-ai/verifiers
 Project-URL: Documentation, https://github.com/primeintellect-ai/verifiers

{verifiers-0.2.1.dev3 → verifiers-0.2.1.dev4}/tests/v1/conftest.py

@@ -172,6 +172,7 @@ def _eval_config(
     harness_overrides: dict | None = None,
     pool: dict | None = None,
     model: str | None = None,
+    reasoning_effort: str | None = None,
 ) -> EvalConfig:
     """Build the smallest `EvalConfig` that still exercises the path, shared by the in-process
     (`run_v1`) and env-server (`run_v1_server`) fixtures. `taskset_overrides` / `harness_overrides`
@@ -192,7 +193,11 @@ def _eval_config(
         num_rollouts=n,
         max_turns=max_turns,
         max_output_tokens=max_tokens,
-        sampling={"max_tokens": max_tokens, "temperature": 0},
+        sampling={
+            "max_tokens": max_tokens,
+            "temperature": 0,
+            "reasoning_effort": reasoning_effort,
+        },
         timeout={"rollout": rollout_timeout, "scoring": 60},
         retries={"rollout": {"max_retries": 2, "include": ["ProviderError"]}},
         rich=False,

{verifiers-0.2.1.dev3 → verifiers-0.2.1.dev4}/tests/v1/test_e2e.py

@@ -145,7 +145,8 @@ async def test_tool_response_image(run_v1, tmp_path):
         "tool-response-image-v1",
         harness="null",
         harness_overrides={"runtime": {"type": "subprocess"}},
-        model="qwen/qwen3-vl-8b-instruct",
+        model="openai/gpt-5.6-luna",
+        reasoning_effort="none",
         output_dir=tmp_path,
         max_turns=4,
     )