Metadata-Version: 2.4
Name: opentrain-verify
Version: 0.1.0
Summary: Human verification and reward functions for RL training loops — the thin OpenTrain verify client.
Project-URL: Homepage, https://www.opentrain.ai
Project-URL: Repository, https://github.com/OpenTrain-AI/OpenTrain-Platform-Monorepo
Author-email: OpenTrain <support@opentrain.ai>
License-Expression: Apache-2.0
License-File: LICENSE
Keywords: human-reward,opentrain,prime-intellect,rlvr,verifiers
Classifier: Development Status :: 4 - Beta
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Typing :: Typed
Requires-Python: >=3.10
Requires-Dist: opentrain-ai>=0.1.0
Description-Content-Type: text/markdown

# opentrain-verify

The thin OpenTrain verify client for the RL crowd: human verification
batches, the `opentrain_human_reward()` function for `verifiers` Rubrics,
Prime Intellect Hub rollout pushes, and offline Certificate v2 checks —
without installing the full SDK surface.

```bash
pip install opentrain-verify
```

```python
from opentrain_verify import OpenTrain, opentrain_human_reward

client = OpenTrain()  # OPENTRAIN_API_TOKEN
reward_fn = opentrain_human_reward(client, rubric=["The answer is correct."])
# rubric = vf.Rubric(funcs=[reward_fn, ...], weights=[1.0, ...])
```

Everything here re-exports from the `opentrain-ai` package (imports as `opentrain`) — see its README
for the full API.
