推理端点(专用)文档

支持的 Transformer 及 Diffusers 任务

Hugging Face's logo
加入 Hugging Face 社区

并获取增强型文档体验

开始

支持可逆变换 & Diffusers 任务

推理端点为以下库中的机器学习任务提供开箱即用的支持

  • Transformer
  • 句子变换器
  • Diffusers(用于文本到图像任务)

下面是 Hugging Face 管理的推理端点支持的任务表。这些任务不需要任何形式的代码或 “自定义容器” 即可部署端点。如果您想自定义以下任何任务,或想编写自己的自定义任务,请参阅 “创建您自己的推理处理器” 部分以获取更多信息。

以下大部分任务使用 pipeline 对象,更多有关可以发送到端点附加参数的信息请访问 此处

任务 框架 开箱即用支持
文本到图像 Diffusers
文本分类 Transformer
零样本分类 Transformer
标记分类 Transformer
问答 Transformer
填空 Transformer
摘要 Transformer
翻译 Transformer
文本到文本生成 Transformer
文本生成 Transformer
特征提取 Transformer
句子嵌入 句子转换器
句子相似度 句子转换器
排序 句子转换器
图像分类 Transformer
自动语音识别 Transformer
音频分类 Transformer
目标检测 Transformer
图像分割 Transformer
表格问答 Transformer
对话式 Transformer
自定义 自定义
视觉问答 Transformer
零样本图像分类 Transformer

示例请求负载

查看以下请求示例了解一些任务

自定义处理程序

{
  "inputs": "This is a sample input",
  "moreData": 1,
  "customTask": true
}

文本分类

有关更多参数,请参阅此参考

对单个文本进行分类

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

对文本对进行分类

{
  "inputs": {
    "text": "This sound track was beautiful!",
    "text_pair": "It paints the scenery in your mind so well I would recomend it even to people who hate vid. game music!"
  } 
}

零样本分类

有关其他参数,请参阅此参考文献

{
  "inputs": "Hi, I recently bought a device from your company but it is not working as advertised and I would like to get reimbursed!",
  "parameters": {
    "candidate_labels": ["refund", "legal", "faq"]
  }
}

标记分类

有关其他参数,请参阅此参考文献

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

问答

有关其他参数,请参阅此参考文献

{
  "inputs": {
    "question": "What is used for inference?",
    "context": "My Name is Philipp and I live in Nuremberg. This model is used with sagemaker for inference."
  }
}

填充掩码

有关其他参数,请参阅此参考文献

{
  "inputs": "This sound track was <mask>! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

摘要

关于其他参数,请参阅此参考

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

翻译

关于其他参数,请参阅此参考

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

文本到文本生成

关于其他参数,请参阅此参考

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

文本生成

关于其他参数,请参阅此参考

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

特征提取

有关其他参数,请参见此参考

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

句子嵌入

如果使用TEI容器,请参考此处获取有关参数的更多信息。

{
  "inputs": "This sound track was beautiful! It paints the scenery in your mind so well I would recomend it
  even to people who hate vid. game music!"
}

句子相似度

{
  "inputs": {
    "sentences": ["This sound track was beautiful!", "It paints the scenery in your mind so well"],
    "source_sentence": "What a wonderful day to listen to music"
  }
}

排序

{
  "inputs": ["This sound track was beautiful!", "It paints the scenery in your mind so well"]
}

图片分类

图片分类可以直接接收来自图片的 JSON 负载或二进制数据。

JSON

{
  "inputs": "/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgI"
}

二进制

curl --request POST \
  --url https://{ENDPOINT}/ \
  --header 'Content-Type: image/jpg' \
  --header 'Authorization: Bearer {HF_TOKEN}' \
  --data-binary '@test.jpg'

自动语音识别

自动语音识别可以直接接收来自音频的 JSON 负载或二进制数据。有关其他参数,请参阅此参考

JSON

{
  "inputs": "/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgI"
}

二进制

curl --request POST \
  --url https://{ENDPOINT}/ \
  --header 'Content-Type: audio/x-flac' \
  --header 'Authorization: Bearer {HF_TOKEN}' \
  --data-binary '@sample.flac'

音频分类

音频分类可以直接接收来自音频的 JSON 负载或二进制数据。有关其他参数,请参阅此 参考

JSON

{
  "inputs": "/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgI"
}

二进制

curl --request POST \
  --url https://{ENDPOINT}/ \
  --header 'Content-Type: audio/x-flac' \
  --header 'Authorization: Bearer {HF_TOKEN}' \
  --data-binary '@sample.flac'

目标检测

目标检测可以直接接收来自图片的 JSON 负载或二进制数据。有关其他参数,请参阅此 参考

JSON

{
  "inputs": "/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgI"
}

二进制

curl --request POST \
  --url https://{ENDPOINT}/ \
  --header 'Content-Type: image/jpg' \
  --header 'Authorization: Bearer {HF_TOKEN}' \
  --data-binary '@test.jpg'

图像分割

图像分割可以从直接从image接收json有效载荷或二进制数据。对于其他参数,请参阅此参考资料

JSON

{
  "inputs": "/9j/4AAQSkZJRgABAQEBLAEsAAD/2wBDAAMCAgI"
}

二进制

curl --request POST \
  --url https://{ENDPOINT}/ \
  --header 'Content-Type: image/jpg' \
  --header 'Authorization: Bearer {HF_TOKEN}' \
  --data-binary '@test.jpg'

表格问答

对于其他参数,请参阅此参考资料

{
  "inputs": {
    "query": "How many stars does the transformers repository have?",
    "table": {
      "Repository": ["Transformers", "Datasets", "Tokenizers"],
      "Stars": ["36542", "4512", "3934"],
      "Contributors": ["651", "77", "34"],
      "Programming language": ["Python", "Python", "Rust, Python and NodeJS"]
    }
  }
}

对话

对于其他参数,请参阅此参考资料

{"inputs": [
  {
      "role": "user",
      "content": "Which movie is the best ?"
  },
  {
      "role": "assistant",
      "content": "It's Die Hard for sure."
  },
  {
      "role": "user",
      "content": "Can you explain why?"
  }
]}

文本转图像

{        
  "inputs": "realistic render portrait realistic render portrait of group of flying blue whales towards the moon, intricate, toy, sci - fi, extremely detailed, digital painting, sculpted in zbrush, artstation, concept art, smooth, sharp focus, illustration, chiaroscuro lighting, golden ratio, incredible art by artgerm and greg rutkowski and alphonse mucha and simon stalenhag",
}

对于文本到图像模型,请注意,当前您的模型仓库需要一个包含完全权重(即,不仅仅是LoRA)的diffusers模型。

附加参数

您可以添加由 transformers 的 pipelines api 支持的附加参数。

例如,如果您有一个 text-generation 管道,您可以为 repetition_penaltymax_length 提供额外的 generation_kwargs 参数。

{
  "inputs": "Hugging Face, the winner of VentureBeat’s Innovation in Natural Language Process/Understanding Award for 2021, is looking to level the playing field. The team, launched by Clément Delangue and Julien Chaumond in 2016, was recognized for its work in democratizing NLP, the global market value for which is expected to hit $35.1 billion by 2026. This week, Google’s former head of Ethical AI Margaret Mitchell joined the team.",
  "parameters": {
    "repetition_penalty": 4.0,
    "max_length": 128
  }
}
< > 更新于 GitHub