---
dataset_info:
  features:
  - name: audio_path
    dtype: string
  - name: doc_id
    dtype: string
  - name: src_text
    dtype: string
  - name: src_text_system
    dtype: string
  - name: src_lang
    dtype: string
  - name: tgt_lang
    dtype: string
  - name: domain
    dtype: string
  - name: tgt_system
    dtype: string
  - name: tgt_text
    dtype: string
  - name: score
    dtype: float64
  - name: audio
    dtype: audio
  splits:
  - name: train
    num_bytes: 8046291442
    num_examples: 33721
  - name: train_synthetic
    num_bytes: 324856288
    num_examples: 7000
  - name: dev
    num_bytes: 1241951905
    num_examples: 5556
  download_size: 71228386133
  dataset_size: 9613099635
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: train_synthetic
    path: data/train_synthetic-*
  - split: dev
    path: data/dev-*
license: cc-by-nc-nd-4.0
tags:
- quality-estimation
- speech-translation
- speech
size_categories:
- 10K<n<100K
---

# Metrics-Shared Task IWSLT 2026 Train and Dev Set

This dataset contains the train and dev sets for the **Speech Translation Metrics Shared Task** at **IWSLT 2026**. More details about the shared task can be found on the [IWSLT website](https://iwslt.org/2026/metrics).

The dataset is primarily designed for research in speech translation quality estimation. 

## Task Goal

Given a speech sample and a system-generated translation, the goal is to estimate a score that reflects the translation quality.

---

## Dataset Splits

### train
Contains a mix of:
- **IWSLT 2023** human annotations ([details IWSLT 2023](https://aclanthology.org/2023.iwslt-1.1/))
  - Previous and following segments can be inferred from the `doc_id` feature.
  - Human evaluators considered context of one previous and one following segment.
- **WMT 2024** human annotations ([details WMT 2024](https://aclanthology.org/2024.wmt-1.1/))
  - Evaluated on segmented audio. The information on previous/following segments is not available.
- **WMT 2025** human annotations ([details WMT 2025](https://aclanthology.org/2025.wmt-1.22/))
  - Evaluated on segmented audio. The information on previous/following segments is not available.


### train_synthetic
Contains:
- **SpeechQE** data ([details SpeechQE](https://aclanthology.org/2024.emnlp-main.1218/))
  - Based on Common Voice.
  - Automatically annotated (synthetic scores).

### dev
Contains:
- **IWSLT 2025 ACL Talks** human annotations ([details IWSLT 2025](https://aclanthology.org/2025.iwslt-1.44/))
  - Previous and following segments can be inferred from the `doc_id` feature.
  - Human evaluators considered context of one previous and one following segment.

---

## Features

| Column | Type | Description |
|--------|------|-------------|
| `audio` | `Audio` | The speech waveform of the segment |
| `audio_path` | `string` | Path to the audio file |
| `doc_id` | `string` | Unique identifier for the segment/document |
| `src_text` | `string` | Source text  |
| `src_text_system` | `string` | Source text system (e.g. human, ASR model) |
| `src_lang` | `string` | Source language code (e.g., `en`) |
| `tgt_text` | `string` | Target text translation |
| `tgt_lang` | `string` | Target language code (e.g., `de`) |
| `domain` | `string` | Domain or dataset source |
| `tgt_system` | `string` | Target system or model used for translation |
| `score` | `float64` | Human or synthetic evaluation score (0–1) |

---

## Dataset Card Contact 
Maike Züfle [@maikezu](https://huggingface.co/maikezu)