PredictPreTrainModel
UIE extraction is a built-in capability used for general-purpose pre-labeling.
Try it now
Try this API in OpenAPI Explorer, no manual signing needed. Successful calls auto-generate SDK code matching your parameters. Download it with built-in credential security for local usage.Try this API in OpenAPI Explorer, no manual signing needed. Successful calls auto-generate SDK code matching your parameters. Download it with built-in credential security for local usage.
Test
RAM authorization
|
Action |
Access level |
Resource type |
Condition key |
Dependent action |
|
documentautoml:PredictPreTrainModel |
get |
*All Resource
|
None | None |
Request parameters
|
Parameter |
Type |
Required |
Description |
Example |
| Content |
string |
No |
The prediction content, in JSON string format. For UIE extraction, the format is |
{"query":"https://doc-automl-public.oss-cn-hangzhou.aliyuncs.com/demo/extractBill.png", "schema":["姓名", "住址"]} |
| ServiceVersion |
string |
No |
The version of the pre-built capability service. The default value is |
V1 |
| ServiceName |
string |
No |
The name of the pre-built capability service. For UIE extraction, set this parameter to |
FormUIE |
| BinaryToText |
boolean |
No |
Set this parameter to |
false |
| Body |
string |
No |
The prediction content for a Base64-encoded image, in JSON string format. For UIE extraction, the format is |
{"query":"data:image/png;base64,xxxxx","schema":["姓名", "住址"]} |
The BinaryToText parameter is optional.
Provide the image content using either the content parameter for an image URL or the body parameter for base64-encoded image data. When using the body parameter, you must also set BinaryToText to true.
Response elements
|
Element |
Type |
Description |
Example |
|
object |
The response object. |
||
| RequestId |
string |
A unique identifier for the request. |
29413C69-11EF-15CB-BE20-70D318E2F4E9 |
| Code |
integer |
The request status. A value of |
200 |
| Message |
string |
A message describing the request status. If the request fails, this parameter contains the error message. |
success |
| Data |
object |
An object containing the extracted information. |
{ "RequestId": "29413C69-11EF-15CB-BE20-70D318E2F4E9", "Message": "success", "Data": { "score": 1, "data": [ { "prob": 0.9138551354408264, "fieldWordRaw": "xx", "wordInfo": [ { "prob": 0.9138551354408264, "pos": [ { "x": 468, "y": 139 }, { "x": 476, "y": 139 }, { "x": 476, "y": 153 }, { "x": 468, "y": 153 } ], "word": "xx", "charInfo": [ { "prob": 0.99, "pos": [ { "x": 468, "y": 139 }, { "x": 471, "y": 139 }, { "x": 471, "y": 153 }, { "x": 468, "y": 153 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 473, "y": 139 }, { "x": 476, "y": 139 }, { "x": 476, "y": 153 }, { "x": 473, "y": 153 } ], "word": "x" } ] } ], "name": "姓名", "location": [ { "x": 468, "y": 139 }, { "x": 476, "y": 139 }, { "x": 476, "y": 153 }, { "x": 468, "y": 153 } ], "fieldWord": "xx" }, { "prob": 0.9594210982322693, "fieldWordRaw": "xxxxxxxxxxx", "wordInfo": [ { "prob": 0.9594210982322693, "pos": [ { "x": 168, "y": 164 }, { "x": 467, "y": 164 }, { "x": 467, "y": 239 }, { "x": 168, "y": 239 } ], "word": "xxxxxxxxxxx", "charInfo": [ { "prob": 0.99, "pos": [ { "x": 366, "y": 164 }, { "x": 383, "y": 164 }, { "x": 383, "y": 179 }, { "x": 366, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 384, "y": 164 }, { "x": 401, "y": 164 }, { "x": 401, "y": 179 }, { "x": 384, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 402, "y": 164 }, { "x": 414, "y": 164 }, { "x": 414, "y": 179 }, { "x": 402, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 418, "y": 164 }, { "x": 427, "y": 164 }, { "x": 427, "y": 179 }, { "x": 418, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 429, "y": 164 }, { "x": 438, "y": 164 }, { "x": 438, "y": 179 }, { "x": 429, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 439, "y": 164 }, { "x": 448, "y": 164 }, { "x": 448, "y": 179 }, { "x": 439, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 450, "y": 164 }, { "x": 467, "y": 164 }, { "x": 467, "y": 179 }, { "x": 450, "y": 179 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 168, "y": 224 }, { "x": 180, "y": 224 }, { "x": 180, "y": 239 }, { "x": 168, "y": 239 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 180, "y": 224 }, { "x": 192, "y": 224 }, { "x": 192, "y": 239 }, { "x": 180, "y": 239 } ], "word": "x" }, { "prob": 0.99, "pos": [ { "x": 195, "y": 224 }, { "x": 209, "y": 224 }, { "x": 209, "y": 239 }, { "x": 195, "y": 239 } ], "word": "1" }, { "prob": 0.99, "pos": [ { "x": 211, "y": 224 }, { "x": 223, "y": 224 }, { "x": 223, "y": 239 }, { "x": 211, "y": 239 } ], "word": "7" }, { "prob": 0.99, "pos": [ { "x": 224, "y": 224 }, { "x": 236, "y": 224 }, { "x": 236, "y": 239 }, { "x": 224, "y": 239 } ], "word": "2" }, { "prob": 0.99, "pos": [ { "x": 237, "y": 224 }, { "x": 249, "y": 224 }, { "x": 249, "y": 239 }, { "x": 237, "y": 239 } ], "word": "7" }, { "prob": 0.99, "pos": [ { "x": 253, "y": 224 }, { "x": 270, "y": 224 }, { "x": 270, "y": 239 }, { "x": 253, "y": 239 } ], "word": "x" } ] } ], "name": "住址", "location": [ { "x": 168, "y": 164 }, { "x": 467, "y": 164 }, { "x": 467, "y": 239 }, { "x": 168, "y": 239 } ], "fieldWord": "xxxxxxxx" } ], "specificType": "formuie-anti", "classType": "preDefinedServer", "predictFile": "https://xxxx" }, "Code": 200 } |
The Data object returned by the template service prediction API contains the following fields:
score: The overall confidence score for the prediction, ranging from 0.0 to 1.0.
data: An array of detailed prediction results returned by the model.
prob: The confidence score for a specific extracted item, ranging from 0.0 to 1.0.
name: The name of the extracted field, which is the key in a key-value pair.
fieldWord: The value of the extracted field.
location: The location of the extracted item on the page, specified as an array of x and y coordinates that form a bounding box. A single coordinate is represented as { "x": 119, "y": 48 }.
wordInfo: Details of the extracted content, including the location of each character.
specificType: The algorithm type for the extraction. For example, formuie-anti indicates Universal Information Extraction (UIE).
classType: The service classification. For example, preDefinedServer indicates a pre-built capability.
predictFile: The prediction file URL, which expires in 60 minutes.
Examples
Success response
JSON format
{
"RequestId": "29413C69-11EF-15CB-BE20-70D318E2F4E9",
"Code": 200,
"Message": "success",
"Data": {
"RequestId": "29413C69-11EF-15CB-BE20-70D318E2F4E9",
"Message": "success",
"Data": {
"score": 1,
"data": [
{
"prob": 0.9138551354408264,
"fieldWordRaw": "xx",
"wordInfo": [
{
"prob": 0.9138551354408264,
"pos": [
{
"x": 468,
"y": 139
},
{
"x": 476,
"y": 139
},
{
"x": 476,
"y": 153
},
{
"x": 468,
"y": 153
}
],
"word": "xx",
"charInfo": [
{
"prob": 0.99,
"pos": [
{
"x": 468,
"y": 139
},
{
"x": 471,
"y": 139
},
{
"x": 471,
"y": 153
},
{
"x": 468,
"y": 153
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 473,
"y": 139
},
{
"x": 476,
"y": 139
},
{
"x": 476,
"y": 153
},
{
"x": 473,
"y": 153
}
],
"word": "x"
}
]
}
],
"name": "姓名",
"location": [
{
"x": 468,
"y": 139
},
{
"x": 476,
"y": 139
},
{
"x": 476,
"y": 153
},
{
"x": 468,
"y": 153
}
],
"fieldWord": "xx"
},
{
"prob": 0.9594210982322693,
"fieldWordRaw": "xxxxxxxxxxx",
"wordInfo": [
{
"prob": 0.9594210982322693,
"pos": [
{
"x": 168,
"y": 164
},
{
"x": 467,
"y": 164
},
{
"x": 467,
"y": 239
},
{
"x": 168,
"y": 239
}
],
"word": "xxxxxxxxxxx",
"charInfo": [
{
"prob": 0.99,
"pos": [
{
"x": 366,
"y": 164
},
{
"x": 383,
"y": 164
},
{
"x": 383,
"y": 179
},
{
"x": 366,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 384,
"y": 164
},
{
"x": 401,
"y": 164
},
{
"x": 401,
"y": 179
},
{
"x": 384,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 402,
"y": 164
},
{
"x": 414,
"y": 164
},
{
"x": 414,
"y": 179
},
{
"x": 402,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 418,
"y": 164
},
{
"x": 427,
"y": 164
},
{
"x": 427,
"y": 179
},
{
"x": 418,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 429,
"y": 164
},
{
"x": 438,
"y": 164
},
{
"x": 438,
"y": 179
},
{
"x": 429,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 439,
"y": 164
},
{
"x": 448,
"y": 164
},
{
"x": 448,
"y": 179
},
{
"x": 439,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 450,
"y": 164
},
{
"x": 467,
"y": 164
},
{
"x": 467,
"y": 179
},
{
"x": 450,
"y": 179
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 168,
"y": 224
},
{
"x": 180,
"y": 224
},
{
"x": 180,
"y": 239
},
{
"x": 168,
"y": 239
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 180,
"y": 224
},
{
"x": 192,
"y": 224
},
{
"x": 192,
"y": 239
},
{
"x": 180,
"y": 239
}
],
"word": "x"
},
{
"prob": 0.99,
"pos": [
{
"x": 195,
"y": 224
},
{
"x": 209,
"y": 224
},
{
"x": 209,
"y": 239
},
{
"x": 195,
"y": 239
}
],
"word": "1"
},
{
"prob": 0.99,
"pos": [
{
"x": 211,
"y": 224
},
{
"x": 223,
"y": 224
},
{
"x": 223,
"y": 239
},
{
"x": 211,
"y": 239
}
],
"word": "7"
},
{
"prob": 0.99,
"pos": [
{
"x": 224,
"y": 224
},
{
"x": 236,
"y": 224
},
{
"x": 236,
"y": 239
},
{
"x": 224,
"y": 239
}
],
"word": "2"
},
{
"prob": 0.99,
"pos": [
{
"x": 237,
"y": 224
},
{
"x": 249,
"y": 224
},
{
"x": 249,
"y": 239
},
{
"x": 237,
"y": 239
}
],
"word": "7"
},
{
"prob": 0.99,
"pos": [
{
"x": 253,
"y": 224
},
{
"x": 270,
"y": 224
},
{
"x": 270,
"y": 239
},
{
"x": 253,
"y": 239
}
],
"word": "x"
}
]
}
],
"name": "住址",
"location": [
{
"x": 168,
"y": 164
},
{
"x": 467,
"y": 164
},
{
"x": 467,
"y": 239
},
{
"x": 168,
"y": 239
}
],
"fieldWord": "xxxxxxxx"
}
],
"specificType": "formuie-anti",
"classType": "preDefinedServer",
"predictFile": "https://xxxx"
},
"Code": 200
}
}Error codes
|
HTTP status code |
Error code |
Error message |
Description |
|---|---|---|---|
| 200 | 21002 | 模板预测超时 | |
| 200 | 21003 | 模版预测失败 | |
| 200 | 10001 | 参数出错 | |
| 200 | 16001 | 未找到可预测的模型 | |
| 200 | 13018 | 未找到模型信息 | |
| 200 | 16004 | 指定模型不存在 | |
| 200 | 23002 | 获取资源HTTP异常 | |
| 200 | 11002 | 账号没有开通服务 | |
| 200 | 19999 | 未知异常 | |
| 200 | 50008 | 文件过大 | |
| 200 | 50007 | 文件格式不支持 | |
| 200 | 10005 | 服务不存在 |
See Error CodesError Codes for a complete list.
Release notes
See Release NotesRelease Notes for a complete list.