Download example_adapter.py from vynly/nnai-bench: direct link, hf CLI and curl.
- Browser
- Download file 406 Bytes
-
https://huggingface.co/datasets/vynly/nnai-bench/resolve/main/example_adapter.py
- Command line
-
hf download hf://datasets/vynly/nnai-bench/example_adapter.py
-
curl -L -o example_adapter.py https://huggingface.co/datasets/vynly/nnai-bench/resolve/main/example_adapter.py
406 Bytes
| """Adapter example: fixed A control. Replace predict() with your non-neural model.""" | |
| import json | |
| import sys | |
| def predict(question): | |
| # Only prompt / choices / task metadata are provided. Never read gold labels. | |
| return "A" | |
| if __name__ == "__main__": | |
| for line in sys.stdin: | |
| question = json.loads(line) | |
| print(json.dumps({"id": question["id"], "prediction": predict(question)})) | |