The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
duvo-eye-1 — benchmark evidence
Per-sample predictions and eval harness for duvoai/duvo-eye-1 public-benchmark results. Every claimed number can be independently rescored from the .predictions.jsonl files (one line per sample: raw model output, parsed point, ground truth, hit/miss).
Protocol: single-shot, temperature 0, max_tokens 64, thinking disabled, guided JSON {"x": int, "y": int} in [0,1000] scaled to original pixels. Harness: bench_eval.py (included).
| Benchmark | duvo-eye-1 | Holo-3.1-35B-A3B (base, same harness) | Files |
|---|---|---|---|
| OSWorld-G (510 standard subset) | 78.0 (398/510) | 64.9 | osworld-g/ |
| ScreenSpot-Pro (max_pixels 8M) | 72.9 | 56.1 | screenspot-pro/ (upload in progress) |
Refusal note: guided decoding always emits a coordinate, so the model cannot earn OSWorld-G refusal credit; on the full 564 set with refusals scored as misses, duvo-eye-1 = 70.6.
- Downloads last month
- 69