Rapidata/text-2-audio-human-preference-benchmark
收藏Hugging Face2025-11-27 更新2026-02-07 收录
下载链接：
https://hf-mirror.com/datasets/Rapidata/text-2-audio-human-preference-benchmark
下载链接
链接失效反馈官方服务：
资源简介：
---
dataset_info:
features:
- name: audio1
dtype: string
- name: audio2
dtype: string
- name: model1
dtype: string
- name: model2
dtype: string
- name: Friendly_weighted_score_audio_1
dtype: float64
- name: Friendly_weighted_score_audio_2
dtype: float64
- name: Friendly_detailedResults
list:
- name: userDetails
struct:
- name: age
dtype: string
- name: country
dtype: string
- name: gender
dtype: string
- name: language
dtype: string
- name: occupation
dtype: string
- name: userScores
struct:
- name: audio_on
dtype: float64
- name: global
dtype: float64
- name: votedFor
dtype: string
- name: Natural_weighted_score_audio_1
dtype: float64
- name: Natural_weighted_score_audio_2
dtype: float64
- name: Natural_detailedResults
list:
- name: userDetails
struct:
- name: age
dtype: string
- name: country
dtype: string
- name: gender
dtype: string
- name: language
dtype: string
- name: occupation
dtype: string
- name: userScores
struct:
- name: audio_on
dtype: float64
- name: global
dtype: float64
- name: votedFor
dtype: string
splits:
- name: train
num_bytes: 4487816
num_examples: 4269
download_size: 407878
dataset_size: 4487816
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
task_categories:
- text-to-speech
pretty_name: Text to Audio Human Preference Benchmark
tags:
- t2a
- text-2-audio
- minimax
- google-gemini-2.5-pro-tts
- elevenlabs
- openai
- openai-gpt-4o-tts
- openai-gpt-4o-mini-tts
---
<style>
.link-container {
padding: 10px;
text-align: center;
border: 1px solid #000000;
border-radius: .25rem;
}
.image{
margin:0 auto;
}
</style>
# Text to Audio Human Benchmark
<a href="https://www.rapidata.ai">
<img src="https://cdn-uploads.huggingface.co/production/uploads/66f5624c42b853e73e0738eb/jfxR79bOztqaC6_yNNnGU.jpeg" width="400" alt="Dataset visualization">
</a>
In this dataset, ~32k human responses collected in less than 1h using the [Rapidata Python API](https://docs.rapidata.ai/mri/), accessible to anyone and ideal for large scale evaluation.
The annotators were asked **Which voice is more friendly?** and **Which voice sounds more natural?** respectively.
<br />
<a href="https://app.rapidata.ai/mri/benchmarks/68ebcd86bd00d9b672b98995">
<div class="link-container">
<div>
Check out the Benchmark!
</div>
<img class="image" src="https://cdn-uploads.huggingface.co/production/uploads/672b7d79fd1e92e3c3567435/Fp0fv_5cW6toqXwSozDEk.png" alt="Audio Benchmark">
</div>
</a>
提供机构：
Rapidata
