3. 返回參數(shù)

Request:

curl "https://api.openai.com/v1/chat/completions" \

  -H "Content-Type: application/json" \

  // 這里要替換成你的API Key

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -d '{

    "model": "gpt-3.5-turbo",

    "messages": [{"role": "system", "content": "You are a helpful assistant."}, {"role": "user", "content": "Hello!"}]

  }'

Response:

{

    // GPT響應(yīng)的唯一表示

    "id": "chatcmpl-123",

    // 表示接口類型

    "object": "chat.completion",

    // 創(chuàng)建的unix時(shí)間戳

    "created": 1677652288,

    // 模型生成的內(nèi)容

    "choices": [{

        // 序號(hào)

        "index": 0,

        // 生成內(nèi)容

        "message": {

            "role": "assistant",

            "content": "\n\nHello there, how may I assist you today?",

        },

        /* 

          停止生成原因：

              stop，正常結(jié)束或者遇到停止序列而停止

              length，生成內(nèi)容達(dá)到了最大長(zhǎng)度限制

              function_call，需要調(diào)用函數(shù)而停止

              content_filter，內(nèi)容被過濾而停止

        */

        "finish_reason": "stop"

    }],

    // 統(tǒng)計(jì)數(shù)據(jù)

    "usage": {

        // 傳入數(shù)據(jù)消耗的token數(shù)

        "prompt_tokens": 9,

        // GPT生成內(nèi)容的Token數(shù)

        "completion_tokens": 12,

        // 所有的Token數(shù)

        "total_tokens": 21

    }

}

4. 代碼示例

Python:

Demo 地址：python-gpt-course/course/openai_api/chat.py

運(yùn)行示例：python course openai_api chat

Golang:

Demo 地址：golang-gpt-course/internal/openai_api/chat.go

運(yùn)行示例：./start openai_api chat

Completions（單輪對(duì)話）

1. 請(qǐng)求地址

POST https://api.openai.com/v1/completions

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request:

curl "https://api.openai.com/v1/completions" \

  -H "Content-Type: application/json" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -d '{

    "model": "text-davinci-003",

    "prompt": "Say this is a test",

    "max_tokens": 7,

    "temperature": 0

  }'

Reponse:

{

    "id": "cmpl-uqkvlQyYK7bGYrRHQ0eXlWi7",

    "object": "text_completion",

    "created": 1589478378,

    "model": "text-davinci-003",

    "choices": [{

        "text": "\n\nThis is indeed a test",

        "index": 0,

        "logprobs": null,

        "finish_reason": "length"

    }],

    "usage": {

        "prompt_tokens": 5,

        "completion_tokens": 7,

        "total_tokens": 12

    }

}

4. 代碼示例

Python:

Demo 地址：python-gpt-course/course/openai_api/completion.py

運(yùn)行示例：python course openai_api completion

Golang:

Demo 地址：golang-gpt-course/internal/openai_api/completion.go

運(yùn)行示例：./start openai_api completion

二、私有化模型構(gòu)建類

這類是用于構(gòu)建私有化模型的相關(guān)接口。私有化模型構(gòu)建分為兩種方式：Embeddings、Fine-tunes。兩個(gè)模型這里簡(jiǎn)單的介紹下，：

Embeddings，以改變上下文的方式來打造私有化模型，該方式不對(duì) GPT 本身的模型進(jìn)行調(diào)整。

Fine-tunes，基于 GPT 的模型再進(jìn)行額外的訓(xùn)練，會(huì)對(duì) GPT 模型本身的參數(shù)進(jìn)行微調(diào)。Fine-tunes 對(duì)模型訓(xùn)練的專業(yè)技能要求更高

PS：在后續(xù)的課程中還會(huì)詳細(xì)講解這兩種方式，所以你可以簡(jiǎn)單看下相關(guān)接口，也可以等到學(xué)習(xí)私有化模型構(gòu)建的時(shí)候再回過頭來看

Embeddings

1. 請(qǐng)求地址

POST https://api.openai.com/v1/embeddings

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/embeddings" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -H "Content-Type: application/json" \

  -d '{

    "input": "The food was delicious and the waiter...",

    "model": "text-embedding-ada-002"

  }

Response

{

    "object": "list",

    "data": [

        {

            "object": "embedding",

            // 向量化數(shù)據(jù)

            "embedding": [

                0.0023064255,

                -0.009327292,

                .... (1536 floats total for ada-002)

                -0.0028842222,

            ],

            "index": 0

        }

    ],

    "model": "text-embedding-ada-002",

    "usage": {

        "prompt_tokens": 8,

        "total_tokens": 8

    }

}

4. 代碼示例

Python:

Demo 地址：python-gpt-course/course/openai_api/embedding.py

運(yùn)行示例：python course openai_api embedding

Golang:

Demo 地址：golang-gpt-course/internal/openai_api/embedding.go

運(yùn)行示例：./start openai_api embedding

Fine-tunings

由于 Fine-tunings 的接口比較多，這里只會(huì)詳細(xì)介紹創(chuàng)建 Fine-tunning 的接口，其他會(huì)簡(jiǎn)要介紹。大多數(shù)場(chǎng)景下，都會(huì)“離線”完成模型訓(xùn)練，然后“在線”實(shí)時(shí)調(diào)用訓(xùn)練后的模型。而“離線”訓(xùn)練更多的是借助 OpenAI 的官方工具，具體的在后續(xù)課程中我們會(huì)進(jìn)一步講解。

另外，F(xiàn)ine-tunings 的使用需要有一定的模型訓(xùn)練的技術(shù)基礎(chǔ)以及經(jīng)驗(yàn)要求，比如你要懂：訓(xùn)練集、測(cè)試集、驗(yàn)證集、損失函數(shù)、梯度下降等等。其實(shí)對(duì)于大多數(shù)場(chǎng)景來說，Embeddings 的方式會(huì)更加容易落地。

1. 請(qǐng)求地址

POST https://api.openai.com/v1/fine-tunes

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/fine-tunes" \

  -H "Content-Type: application/json" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -d '{

    // 訓(xùn)練集的FileID

    "training_file": "file-XGinujblHPwGLSztz8cPS8XY"

  }'

Response

{

      // 訓(xùn)練模型的ID

      "id": "ft-AF1WoRqd3aJAHsqc9NY7iL8F",

      "object": "fine-tune",

      "model": "curie",

      "created_at": 1614807352,

      // 模型訓(xùn)練過程中的一些事件

      "events": [

          {

              "object": "fine-tune-event",

              "created_at": 1614807352,

              "level": "info",

              "message": "Job enqueued. Waiting for jobs ahead to complete. Queue number: 0."

          }

      ],

      "fine_tuned_model": null,

      "hyperparams": {

          "batch_size": 4,

          "learning_rate_multiplier": 0.1,

          "n_epochs": 4,

          "prompt_loss_weight": 0.1,

      },

      "organization_id": "org-...",

      "result_files": [],

      "status": "pending",

      "validation_files": [],

      "training_files": [

          {

              "id": "file-XGinujblHPwGLSztz8cPS8XY",

              "object": "file",

              "bytes": 1547276,

              "created_at": 1610062281,

              "filename": "my-data-train.jsonl",

              "purpose": "fine-tune-train"

          }

      ],

      "updated_at": 1614807352,

}

4. 其他接口

獲取 Fine-tunes 訓(xùn)練的模型列表

GET https://api.openai.com/v1/fine-tunes

獲取某一個(gè)模型的詳細(xì)信息

GET https://api.openai.com/v1/fine-tunes/{fine_tune_id}

取消正在訓(xùn)練的模型

POST https://api.openai.com/v1/fine-tunes/{fine_tune_id}/cancel

獲取某一個(gè)模型訓(xùn)練過程中的事件

GET https://api.openai.com/v1/fine-tunes/{fine_tune_id}/events

刪除一個(gè)模型

DELETE https://api.openai.com/v1/models/{model}

三、通用類

Models（獲取 GPT 可用模型列表）

1. 獲取 Models 列表

請(qǐng)求地址

GET https://api.openai.com/v1/models

Request

curl "https://api.openai.com/v1/models" \

  -H "Authorization: Bearer $OPENAI_API_KEY"

Repsonse

{

    "data": [

        {

            // GPT模型ID

            "id": "model-id-0",

            "object": "model",

            "owned_by": "organization-owner",

            // 權(quán)限

            "permission": [...]

        },

        {

            "id": "model-id-1",

            "object": "model",

            "owned_by": "organization-owner",

            "permission": [...]

        },

        {

            "id": "model-id-2",

            "object": "model",

            "owned_by": "openai",

            "permission": [...]

          },

  ],

  "object": "list"

}

2. 獲取 Models 詳情

請(qǐng)求地址

GET https://api.openai.com/v1/models/{model}

Request

curl "https://api.openai.com/v1/models/text-davinci-003" \

  -H "Authorization: Bearer $OPENAI_API_KEY"

Response

{

    "id": "text-davinci-003",

    "object": "model",

    "owned_by": "openai",

    "permission": [...]

}

Files（上傳文件到 GPT）

主要用于 Fine-tunings 的訓(xùn)練數(shù)據(jù)集、驗(yàn)證數(shù)據(jù)集的上傳

1. 上傳文件

Request

curl "https://api.openai.com/v1/files" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  // 文件用途

  -F purpose="fine-tune" \

  -F file="@mydata.jsonl"

Response

{

    "id": "file-XjGxS3KTG0uNmNOK362iJua3",

    "object": "file",

    "bytes": 140,

    "created_at": 1613779121,

    "filename": "mydata.jsonl",

    "purpose": "fine-tune"

}

2. 其他接口

獲取文件列表

GET https://api.openai.com/v1/files

獲取單個(gè)文件描述信息

GET https://api.openai.com/v1/files/{file_id}

獲取單個(gè)文件內(nèi)容

GET https://api.openai.com/v1/files/{file_id}/content

四、圖片 & 音頻 類

Images

1. 文生圖

文字描述圖，GPT 來生成

1. 請(qǐng)求地址

POST https://api.openai.com/v1/images/generations

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/images/generations" \

  -H "Content-Type: application/json" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -d '{

    "prompt": "A cute baby sea otter",

    "n": 2,

    "size": "1024x1024"

  }'

Response

{

    "created": 1589478378,

    "data": [

        {

            "url": "https://..."

        },

        {

            "url": "https://..."

        }

    ]

}

2. 編輯圖

上傳圖片，并描述修改方式，GPT 生成新圖

1. 請(qǐng)求地址

POST https://api.openai.com/v1/images/edits

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/images/edits" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -F image="@otter.png" \

  -F mask="@mask.png" \

  -F prompt="A cute baby sea otter wearing a beret" \

  -F n=2 \

  -F size="1024x1024"

Response

{

    "created": 1589478378,

    "data": [

        {

            "url": "https://..."

        },

        {

            "url": "https://..."

        }

    ]

}

3. 圖變體

4. 請(qǐng)求地址

POST https://api.openai.com/v1/images/variations

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/images/variations" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -F image="@otter.png" \

  -F n=2 \

  -F size="1024x1024"

Response

{

    "created": 1589478378,

    "data": [

        {

            "url": "https://..."

        },

        {

            "url": "https://..."

        }

    ]

}

Audio（語音轉(zhuǎn)文字）

將語音轉(zhuǎn)為文字

1. 請(qǐng)求地址

POST https://api.openai.com/v1/audio/transcriptions

2. 請(qǐng)求參數(shù)

3. 返回參數(shù)

Request

curl "https://api.openai.com/v1/audio/transcriptions" \

  -H "Authorization: Bearer $OPENAI_API_KEY" \

  -H "Content-Type: multipart/form-data" \

  -F file="@/path/to/file/audio.mp3" \

  -F model="whisper-1"

Response

{

    "text": "Imagine the wildest idea that you've ever had, and you're curious about how it might scale to something that's a 100, a 1,000 times bigger. This is a place where you can get to do that."

}

本文章轉(zhuǎn)載微信公眾號(hào)@AI老張

上一篇：

開放數(shù)據(jù)和API接口采用研究——基于美國(guó)銀行業(yè)的分析

下一篇：

接口大師v3.10.0 零代碼/低代碼，把內(nèi)部接口變成OpenAPI開放接口的妙招
#你可能也喜歡這些API文章!

我們有何不同？

API服務(wù)商零注冊(cè)

多API并行試用

數(shù)據(jù)驅(qū)動(dòng)選型，提升決策效率

查看全部API→
??

熱門場(chǎng)景實(shí)測(cè)，選對(duì)API

#AI文本生成大模型API

對(duì)比大模型API的內(nèi)容創(chuàng)意新穎性、情感共鳴力、商業(yè)轉(zhuǎn)化潛力

25個(gè)渠道
一鍵對(duì)比試用API 限時(shí)免費(fèi)

#AI深度推理大模型API

對(duì)比大模型API的邏輯推理準(zhǔn)確性、分析深度、可視化建議合理性

10個(gè)渠道
一鍵對(duì)比試用API 限時(shí)免費(fèi)