---
dataset_info:
- config_name: corpus
  features:
  - name: id
    dtype: string
  - name: text
    dtype: string
  - name: title
    dtype: string
  splits:
  - name: test
    num_bytes: 191
    num_examples: 2
  download_size: 1965
  dataset_size: 191
- config_name: qrels
  features:
  - name: query-id
    dtype: string
  - name: corpus-id
    dtype: string
  - name: score
    dtype: int32
  splits:
  - name: test
    num_bytes: 76
    num_examples: 4
  download_size: 1640
  dataset_size: 76
- config_name: queries
  features:
  - name: id
    dtype: string
  - name: text
    dtype: string
  splits:
  - name: test
    num_bytes: 136
    num_examples: 2
  download_size: 1481
  dataset_size: 136
configs:
- config_name: corpus
  data_files:
  - split: test
    path: corpus/test-*
- config_name: qrels
  data_files:
  - split: test
    path: qrels/test-*
- config_name: queries
  data_files:
  - split: test
    path: queries/test-*
---

# HIKMA Embedding IR — alinf/hikma_ir

Published automatically by the leaderboard backend.

- task type: `ir`
- source date: 2026_10_04
- content digest: `2445a16cf1af07a216d49afe3668ecb3e14e41ab443559f76ef0c85b23c4d65e`
- commit sha: `pending verification`
- published at: `2026-10-05T06:43:14.351574+00:00`
- counts: corpus=2, queries=2, qrels=4

## Configurations

Each configuration exposes a single `test` split:

| config | columns |
| --- | --- |
| `corpus` | `id`, `text`, `title` |
| `queries` | `id`, `text` |
| `qrels` | `query-id`, `corpus-id`, `score` (int32) |