mirror of
https://github.com/modelscope/modelscope.git
synced 2025-12-17 00:37:43 +01:00
1f6b376599
1. refactor maaslib to modelscope 2. fix UT error 3. support pipeline which does not register default model Link: https://code.alibaba-inc.com/Ali-MaaS/MaaS-lib/codereview/8988388
85 lines
3.2 KiB
Markdown
85 lines
3.2 KiB
Markdown
# Pipeline使用教程
|
||
|
||
本文将简单介绍如何使用`pipeline`函数加载模型进行推理。`pipeline`函数支持按照任务类型、模型名称从模型仓库
|
||
拉取模型进行进行推理，当前支持的任务有
|
||
|
||
* 人像抠图 (image-matting)
|
||
* 基于bert的语义情感分析 (bert-sentiment-analysis)
|
||
|
||
本文将从如下方面进行讲解如何使用Pipeline模块：
|
||
* 使用pipeline()函数进行推理
|
||
* 指定特定预处理、特定模型进行推理
|
||
* 不同场景推理任务示例
|
||
|
||
## 环境准备
|
||
详细步骤可以参考 [快速开始](../quick_start.md)
|
||
|
||
## Pipeline基本用法
|
||
|
||
1. pipeline函数支持指定特定任务名称，加载任务默认模型，创建对应Pipeline对象
|
||
执行如下python代码
|
||
```python
|
||
>>> from modelscope.pipelines import pipeline
|
||
>>> img_matting = pipeline(task='image-matting', model='damo/image-matting-person')
|
||
```
|
||
|
||
2. 传入单张图像url进行处理
|
||
``` python
|
||
>>> import cv2
|
||
>>> result = img_matting('http://pai-vision-data-hz.oss-cn-zhangjiakou.aliyuncs.com/data/test/maas/image_matting/test.png')
|
||
>>> cv2.imwrite('result.png', result['output_png'])
|
||
>>> import os.path as osp
|
||
>>> print(f'result file path is {osp.abspath("result.png")}')
|
||
```
|
||
|
||
pipeline对象也支持传入一个列表输入，返回对应输出列表，每个元素对应输入样本的返回结果
|
||
```python
|
||
>>> results = img_matting(
|
||
[
|
||
'http://pai-vision-data-hz.oss-cn-zhangjiakou.aliyuncs.com/data/test/maas/image_matting/test.png',
|
||
'http://pai-vision-data-hz.oss-cn-zhangjiakou.aliyuncs.com/data/test/maas/image_matting/test.png',
|
||
'http://pai-vision-data-hz.oss-cn-zhangjiakou.aliyuncs.com/data/test/maas/image_matting/test.png',
|
||
])
|
||
```
|
||
|
||
如果pipeline对应有一些后处理参数，也支持通过调用时候传入.
|
||
```python
|
||
>>> pipe = pipeline(task_name)
|
||
>>> result = pipe(input, post_process_args)
|
||
```
|
||
|
||
## 指定预处理、模型进行推理
|
||
pipeline函数支持传入实例化的预处理对象、模型对象，从而支持用户在推理过程中定制化预处理、模型。
|
||
下面以文本情感分类为例进行介绍。
|
||
|
||
由于demo模型为EasyNLP提供的模型，首先，安装EasyNLP
|
||
```shell
|
||
pip install https://atp-modelzoo-sh.oss-cn-shanghai.aliyuncs.com/release/package/whl/easynlp-0.0.4-py2.py3-none-any.whl
|
||
```
|
||
|
||
|
||
下载模型文件
|
||
```shell
|
||
wget https://atp-modelzoo-sh.oss-cn-shanghai.aliyuncs.com/release/easynlp_modelzoo/alibaba-pai/bert-base-sst2.zip && unzip bert-base-sst2.zip
|
||
```
|
||
|
||
创建tokenizer和模型
|
||
```python
|
||
>>> from modelscope.models import Model
|
||
>>> from modelscope.preprocessors import SequenceClassificationPreprocessor
|
||
>>> model = Model.from_pretrained('damo/bert-base-sst2')
|
||
>>> tokenizer = SequenceClassificationPreprocessor(
|
||
model.model_dir, first_sequence='sentence', second_sequence=None)
|
||
```
|
||
|
||
使用tokenizer和模型对象创建pipeline
|
||
```python
|
||
>>> from modelscope.pipelines import pipeline
|
||
>>> semantic_cls = pipeline('text-classification', model=model, preprocessor=tokenizer)
|
||
>>> semantic_cls("Hello world!")
|
||
```
|
||
|
||
## 不同场景任务推理示例
|
||
|
||
人像抠图、语义分类建上述两个例子。 其他例子未来添加。
|