happy8825/ecva_instruct_ver2

---
pretty_name: "/hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175 · happy8825/valid_ecva_clean results"
language:
- en
tags:
- video-retrieval
- evaluation
- vllm
---

# /hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175 · happy8825/valid_ecva_clean results

- **Model**: `/hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175`
- **Dataset**: `happy8825/valid_ecva_clean`
- **Generated**: `2025-12-18 10:50:16Z`

## Metrics
| Metric | Value |
| --- | --- |
| Total samples | 924 |
| With GT | 0 |
| Parsed answers | 0 |
| Top-1 accuracy | 0 |
| Recall@5 | 0 |
| MRR | 0 |

The uploaded JSON contains full per-sample predictions produced via `t3_infer_with_vllm.bash`.

### EVQA/ECVA Metrics
| Metric | Value |
| --- | --- |
| EVQA total | 924 |
| EVQA with GT label | 924 |
| EVQA accuracy | 0.569264 |

## Run Summary

```
Saved 924 results to /home/seohyun/vid_understanding/video_retrieval/video_retrieval/output_ecva_instruct/ecva_instruct_ver2.json
Metrics: {
"total": 924,
"with_gt": 0,
"with_parsed_answer": 0,
"top1_acc": 0.0,
"recall_at_5": 0.0,
"mrr": 0.0,
"num_shards": 1,
"shard_index": 0,
"evqa_total": 924,
"evqa_with_gt_label": 924,
"evqa_acc": 0.5692640692640693
}
Pushed ecva_instruct_ver2.jsonl and README to https://huggingface.co/datasets/happy8825/ecva_instruct_ver2
```

pretty_name: "/hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175 · happy8825/valid_ecva_clean 结果集
language:
- 英语
tags:
- 视频检索（video-retrieval)
- 评估（evaluation)
- vllm

# `/hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175 · happy8825/valid_ecva_clean 结果集

- **模型**：`/hub_data4/seohyun/saves/ecva_instruct/full/sft/checkpoint-175`
- **数据集**：`happy8825/valid_ecva_clean`
- **生成时间**：`2025-12-18 10:50:16Z`

## 评估指标
| 指标 | 数值 |
| --- | --- |
| 总样本数 | 924 |
| 含真实标签（Ground Truth）样本数 | 0 |
| 已解析答案数 | 0 |
| Top-1准确率 | 0 |
| 召回率@5（Recall@5） | 0 |
| 平均倒数排名（Mean Reciprocal Rank） | 0 |

本次上传的JSON文件包含通过`t3_infer_with_vllm.bash脚本生成的完整逐样本预测结果。

### EVQA/ECVA 评估指标
| 指标 | 数值 |
| --- | --- |
| EVQA总样本数 | 924 |
| EVQA含真实标签样本数 | 924 |
| EVQA准确率 | 0.569264 |

## 运行摘要

已将924条结果保存至 /home/seohyun/vid_understanding/video_retrieval/video_retrieval/output_ecva_instruct/ecva_instruct_ver2.json
评估指标：{
"total": 924,
"with_gt": 0,
"with_parsed_answer": 0,
"top1_acc": 0.0,
"recall_at_5": 0.0,
"mrr": 0.0,
"num_shards": 1,
"shard_index": 0,
"evqa_total": 924,
"evqa_with_gt_label": 924,
"evqa_acc": 0.5692640692640693
}
已将ecva_instruct_ver2.jsonl与README推送至 https://huggingface.co/datasets/happy8825/ecva_instruct_ver2