TensorFlow deep learning framework model inference pipeline for portrait cutout inference

Overview

In order to enable ModelScope users to quickly and conveniently use various models provided by the platform, a set of fully functional Python library is provided, which includes the implementation of the official ModelScope model. As well as code related to data preprocessing, postprocessing, effect evaluation and other functions required for using these models for inference, finetune and other tasks. It also provides a simple and easy-to-use API and rich usage examples. By calling the library, users can complete tasks such as model reasoning, training, and evaluation by writing just a few lines of code. They can also quickly perform secondary development on this basis to realize their own innovative ideas.

The algorithm model currently provided by the library covers five major AI fields: image, natural language processing, speech, multi-modality, and science, and dozens of application scenario tasks. For specific tasks, please refer to the document: Task introduce.

Deep Learning Framework

ModelScope Library currently supports deep learning frameworks such as Pytorch and Tensorflow. More frameworks will be continuously updated and expanded in the future, so stay tuned! All official models can be used for model inference through the ModelScope Library, and some models can also use the library for training and evaluation. For complete usage information, see the model card for the corresponding model.

Model Inference Pipeline

Model Inference

In deep learning, inference refers to the process by which the model predicts data. When ModelScope performs inference, it uses pipeline to perform necessary operations sequentially. A typical pipeline usually includes three steps: data preprocessing, model forward inference, and data postprocessing.

Pipeline introduction

The pipeline() method is one of the most basic user methods in the ModelScope framework and can be used to quickly perform model inference in various fields. With the pipeline() method, users can easily complete model inference for specific tasks with just one line of code.

The pipeline() method is one of the most basic user methods in the ModelScope framework and can be used to quickly perform model inference in various fields. With the pipeline() method, users can easily complete model inference for specific tasks with just one line of code.

Usage of Pipeline

This article will briefly introduce how to use the pipeline method to load the model for inference. Through the pipeline method, users can easily pull the required model from the model warehouse based on the task type and model name for inference. The main advantage of this method is that it is easy to use and can perform model inference quickly and efficiently. The convenience of the pipeline method is that it provides a direct way to obtain and apply the model without requiring users to understand the specific details of the model, thereby lowering the threshold for using the model. Through the pipeline method, users can focus more on solving problems and

• Environment preparation
• Important parameters
• Basic usage of Pipeline
• Specify preprocessing, Model for inference
• Usage examples of pipeline for different scenario task reasoning

Basic usage of Pipeline

Chinese word segmentation

The pipeline function supports specifying a specific task name. Load the task default model and create the corresponding pipeline object.

Python code

from modelscope.pipelines import pipelineword_segmentation = pipeline('word-segmentation')input_str = '开源技术小栈作者是Tinywan，你知道不？'print(word_segmentation(input_str))

PHP code

<?php $operator = PyCore::import("operator");$builtins = PyCore::import("builtins");$pipeline = PyCore::import('modelscope.pipelines')->pipeline;$word_segmentation = $pipeline("word-segmentation");$input_str = "开源技术小栈作者是Tinywan，你知道不？";PyCore::print($word_segmentation($input_str));

Online conversion tool: https://www.swoole.com/py2php/

Output results

/usr/local/php-8.2.14/bin/php demo.php 2024-03-25 21:41:42,434 - modelscope - INFO - PyTorch version 2.2.1 Found.2024-03-25 21:41:42,434 - modelscope - INFO - Loading ast index from /home/www/.cache/modelscope/ast_indexer2024-03-25 21:41:42,577 - modelscope - INFO - Loading done! Current index file version is 1.13.0, with md5 f54e9d2dceb89a6c989540d66db83a65 and a total number of 972 components indexed2024-03-25 21:41:44,661 - modelscope - WARNING - Model revision not specified, use revision: v1.0.32024-03-25 21:41:44,879 - modelscope - INFO - initiate model from /home/www/.cache/modelscope/hub/damo/nlp_structbert_word-segmentation_chinese-base2024-03-25 21:41:44,879 - modelscope - INFO - initiate model from location /home/www/.cache/modelscope/hub/damo/nlp_structbert_word-segmentation_chinese-base.2024-03-25 21:41:44,880 - modelscope - INFO - initialize model from /home/www/.cache/modelscope/hub/damo/nlp_structbert_word-segmentation_chinese-baseYou are using a model of type bert to instantiate a model of type structbert. This is not supported for all configurations of models and can yield errors.2024-03-25 21:41:48,633 - modelscope - WARNING - No preprocessor field found in cfg.2024-03-25 21:41:48,633 - modelscope - WARNING - No val key and type key found in preprocessor domain of configuration.json file.2024-03-25 21:41:48,633 - modelscope - WARNING - Cannot find available config to build preprocessor at mode inference, current config: {'model_dir': '/home/www/.cache/modelscope/hub/damo/nlp_structbert_word-segmentation_chinese-base'}. trying to build by task and model information.2024-03-25 21:41:48,639 - modelscope - INFO - cuda is not available, using cpu instead.2024-03-25 21:41:48,640 - modelscope - WARNING - No preprocessor field found in cfg.2024-03-25 21:41:48,640 - modelscope - WARNING - No val key and type key found in preprocessor domain of configuration.json file.2024-03-25 21:41:48,640 - modelscope - WARNING - Cannot find available config to build preprocessor at mode inference, current config: {'model_dir': '/home/www/.cache/modelscope/hub/damo/nlp_structbert_word-segmentation_chinese-base', 'sequence_length': 512}. trying to build by task and model information./home/www/anaconda3/envs/tinywan-modelscope/lib/python3.10/site-packages/transformers/modeling_utils.py:962: FutureWarning: The `device` argument is deprecated and will be removed in v5 of Transformers.warnings.warn({'output': ['开源', '技术', '小', '栈', '作者', '是', 'Tinywan', '，', '你', '知道', '不', '？']}

Input multiple samples

The pipeline object also supports passing in multiple sample list inputs and returns the corresponding output list, with each element corresponding to the return result of the input sample. The reasoning method for multiple pieces of text is that the input data is processed individually using an iterator inside the pipeline and then appended to the same return List.

Python code

from modelscope.pipelines import pipelineword_segmentation = pipeline('word-segmentation')inputs =['开源技术小栈作者是Tinywan，你知道不？','webman这个框架不错，建议你看看']print(word_segmentation(inputs))

PHP code

<?php $operator = PyCore::import("operator");$builtins = PyCore::import("builtins");$pipeline = PyCore::import('modelscope.pipelines')->pipeline;$word_segmentation = $pipeline("word-segmentation");$inputs = new PyList(["开源技术小栈作者是Tinywan，你知道不？", "webman这个框架不错，建议你看看"]);PyCore::print($word_segmentation($inputs));

Output

[{'output': ['开源', '技术', '小', '栈', '作者', '是', 'Tinywan', '，', '你', '知道', '不', '？']},{'output': ['webman', '这个', '框架', '不错', '，', '建议', '你', '看看']}]

Batch inference

pipeline's support for batch inference is similar to the above "input multiple texts", the difference is that batch forward inference will be implemented in the model forward process at the batch_size scale specified by the user.

inputs =['今天天气不错，适合出去游玩','这本书很好，建议你看看']# 指定batch_size参数来支持批量推理print(word_segmentation(inputs, batch_size=2))# 输出[{'output': ['今天', '天气', '不错', '，', '适合', '出去', '游玩']}, {'output': ['这', '本', '书', '很', '好', '，', '建议', '你', '看看']}]

Input a data set

from modelscope.msdatasets import MsDatasetfrom modelscope.pipelines import pipelineinputs = ['今天天气不错，适合出去游玩', '这本书很好，建议你看看']dataset = MsDataset.load(inputs, target='sentence')word_segmentation = pipeline('word-segmentation')outputs = word_segmentation(dataset)for o in outputs:print(o)# 输出{'output': ['今天', '天气', '不错', '，', '适合', '出去', '游玩']}{'output': ['这', '本', '书', '很', '好', '，', '建议', '你', '看看']}

Specify preprocessing and model for inference

The pipeline function supports incoming instantiation preprocessing objects and model objects, thereby supporting users to customize preprocessing and models during the inference process.

Creating model objects for inference

Python code

from modelscope.models import Modelfrom modelscope.pipelines import pipelinemodel = Model.from_pretrained('damo/nlp_structbert_word-segmentation_chinese-base')word_segmentation = pipeline('word-segmentation', model=model)inputs =['开源技术小栈作者是Tinywan，你知道不？','webman这个框架不错，建议你看看']print(word_segmentation(inputs))

PHP code

<?php $operator = PyCore::import("operator");$builtins = PyCore::import("builtins");$Model = PyCore::import('modelscope.models')->Model;$pipeline = PyCore::import('modelscope.pipelines')->pipeline;$model = $Model->from_pretrained("damo/nlp_structbert_word-segmentation_chinese-base");$word_segmentation = $pipeline("word-segmentation", model: $model);$inputs = new PyList(["开源技术小栈作者是Tinywan，你知道不？", "webman这个框架不错，建议你看看"]);PyCore::print($word_segmentation($inputs));

Output

[{'output': ['开源', '技术', '小', '栈', '作者', '是', 'Tinywan', '，', '你', '知道', '不', '？']},{'output': ['webman', '这个', '框架', '不错', '，', '建议', '你', '看看']}]

Create preprocessor and model objects for inference

from modelscope.models import Modelfrom modelscope.pipelines import pipelinefrom modelscope.preprocessors import Preprocessor, TokenClassificationTransformersPreprocessormodel = Model.from_pretrained('damo/nlp_structbert_word-segmentation_chinese-base')tokenizer = Preprocessor.from_pretrained(model.model_dir)# Or call the constructor directly: # tokenizer = TokenClassificationTransformersPreprocessor(model.model_dir)word_segmentation = pipeline('word-segmentation', model=model, preprocessor=tokenizer)inputs =['开源技术小栈作者是Tinywan，你知道不？','webman这个框架不错，建议你看看']print(word_segmentation(inputs))[{'output': ['开源', '技术', '小', '栈', '作者', '是', 'Tinywan', '，', '你', '知道', '不', '？']},{'output': ['webman', '这个', '框架', '不错', '，', '建议', '你', '看看']}]

Image

Note:

1. Make sure you have the OpenCV library installed. If it is not installed, you can install it through pip

pip install opencv-python

If not installed, it will prompt: PHP Fatal error: Uncaught PyError: No module named 'cv2' in /home/www/build /ai/demo3.php:4

2. Make sure you have installed the deep learning framework package TensorFlow library

Otherwise it will prompt modelscope.pipelines.cv.image_matting_pipeline requires the TensorFlow library but it was not found in your environment. Checkout the instructions on the installation page: https://www.tensorflow.org/install and follow the ones that match your environment..

The error message indicates that you are trying to use a module named modelscope.pipelines.cv.image_matting_pipeline, which depends on the TensorFlow library. However, the module does not work properly because the necessary TensorFlow dependencies are missing.

You can use the following command to install the latest version of TensorFlow

pip install tensorflow

##Picture

Portrait cutout ('portrait -matting')

Input picture

Picture

Python code

import cv2from modelscope.pipelines import pipelineportrait_matting = pipeline('portrait-matting')result = portrait_matting('https://modelscope.oss-cn-beijing.aliyuncs.com/test/images/image_matting.png')cv2.imwrite('result.png', result['output_img'])

PHP Code tinywan-images.php

<?php $operator = PyCore::import("operator");$builtins = PyCore::import("builtins");$cv2 = PyCore::import('cv2');$pipeline = PyCore::import('modelscope.pipelines')->pipeline;$portrait_matting = $pipeline("portrait-matting");$result = $portrait_matting("https://modelscope.oss-cn-beijing.aliyuncs.com/test/images/image_matting.png");$cv2->imwrite("tinywan_result.png", $result->__getitem__("output_img"));

Load local file image$result = $portrait_matting("./tinywan.png");

Execution result

/usr/local/php-8.2.14/bin/php tinywan-images.php 2024-03-25 22:17:25,630 - modelscope - INFO - PyTorch version 2.2.1 Found.2024-03-25 22:17:25,631 - modelscope - INFO - TensorFlow version 2.16.1 Found.2024-03-25 22:17:25,631 - modelscope - INFO - Loading ast index from /home/www/.cache/modelscope/ast_indexer2024-03-25 22:17:25,668 - modelscope - INFO - Loading done! Current index file version is 1.13.0, with md5 f54e9d2dceb89a6c989540d66db83a65 and a total number of 972 components indexed2024-03-25 22:17:26,990 - modelscope - WARNING - Model revision not specified, use revision: v1.0.02024-03-25 22:17:27.623085: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.2024-03-25 22:17:27.678592: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.2024-03-25 22:17:28.551510: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT2024-03-25 22:17:29,206 - modelscope - INFO - initiate model from /home/www/.cache/modelscope/hub/damo/cv_unet_image-matting2024-03-25 22:17:29,206 - modelscope - INFO - initiate model from location /home/www/.cache/modelscope/hub/damo/cv_unet_image-matting.2024-03-25 22:17:29,209 - modelscope - WARNING - No preprocessor field found in cfg.2024-03-25 22:17:29,210 - modelscope - WARNING - No val key and type key found in preprocessor domain of configuration.json file.2024-03-25 22:17:29,210 - modelscope - WARNING - Cannot find available config to build preprocessor at mode inference, current config: {'model_dir': '/home/www/.cache/modelscope/hub/damo/cv_unet_image-matting'}. trying to build by task and model information.2024-03-25 22:17:29,210 - modelscope - WARNING - Find task: portrait-matting, model type: None. Insufficient information to build preprocessor, skip building preprocessorWARNING:tensorflow:From /home/www/anaconda3/envs/tinywan-modelscope/lib/python3.10/site-packages/modelscope/utils/device.py:60: is_gpu_available (from tensorflow.python.framework.test_util) is deprecated and will be removed in a future version.Instructions for updating:Use `tf.config.list_physical_devices('GPU')` instead.2024-03-25 22:17:29,213 - modelscope - INFO - loading model from /home/www/.cache/modelscope/hub/damo/cv_unet_image-matting/tf_graph.pbWARNING:tensorflow:From /home/www/anaconda3/envs/tinywan-modelscope/lib/python3.10/site-packages/modelscope/pipelines/cv/image_matting_pipeline.py:45: FastGFile.__init__ (from tensorflow.python.platform.gfile) is deprecated and will be removed in a future version.Instructions for updating:Use tf.gfile.GFile.2024-03-25 22:17:29,745 - modelscope - INFO - load model done

Output image

picture

The above is the detailed content of TensorFlow deep learning framework model inference pipeline for portrait cutout inference. For more information, please follow other related articles on the PHP Chinese website!