| id | 347af48c-4d60-47e3-8349-a0139f9ae88c |
|---|---|
| dataset_name | zh-zhihu-answer |
| version_name | 1.1.0-dedup |
| level | L2 |
| parent_version_id | 3660e38b-81fc-410b-90a8-5ce884ee12d0 |
| owner | mhjiang |
| domain | language |
| uri | /inspire/hdd/project/demo/zh-zhihu/dedup |
| format | — |
| item_count | 70000 |
| token_count | — |
| byte_size | — |
| produced_by | filter_dedup |
| manifest_digest | — |
| report_uri | — |
| status | registered |
| metadata_json | {} |
| created_by | — |
| created_at | 2026-07-16 13:40:11 |
| updated_at | 2026-07-16 13:40:11 |
| revision | 1 |
无下游版本
| benchmark_name | metrics_json | model_id | model_version | external_run_ref | status | exposure_status | created_at |
|---|---|---|---|---|---|---|---|
| mmlu | {"accuracy": 0.618} |
qwen3-8b | — | https://eval.platform/runs/2002 | succeeded | unchecked | 2026-07-16 13:40 |
| version_name | benchmark_name | metrics_json | model_id |
|---|---|---|---|
| 1.1.0-dedup (当前) | mmlu | {"accuracy": 0.618} |
qwen3-8b |
| 1.0.0 | mmlu | {"accuracy": 0.512} |
qwen3-8b |