PyPI 官网下载 | fastestimator-nightly-1.1.3.dev202103291252.tar.gz

# FastEstimator [](LICENSE) [](http://jenkins.fastestimator.org:8080/job/fastestimator/job/fastestimator/job/master/) [](http://jenkins.fastestimator.org:8080/job/nightly/) [](https://www.codacy.com/gh/fastestimator/fastestimator/dashboard?utm_source=github.com&utm_medium=referral&utm_content=fastestimator/fastestimator&utm_campaign=Badge_Grade) [](https://www.codacy.com/gh/fastestimator/fastestimator/dashboard?utm_source=github.com&utm_medium=referral&utm_content=fastestimator/fastestimator&utm_campaign=Badge_Coverage) [](https://pypi.org/project/fastestimator/) [](https://pypistats.org/packages/fastestimator) [](https://pypistats.org/packages/fastestimator-nightly) FastEstimator is a high-level deep learning library built on TensorFlow2 and PyTorch. With the help of FastEstimator, you can easily build a high-performance deep learning model and run it anywhere. :wink: ## Prerequisites: * Python >= 3.6 * TensorFlow == 2.4.1 * PyTorch >= 1.7.1 ## Installation: ### 1. Install Dependencies: * Install TensorFlow [here](https://www.tensorflow.org/install) * Install PyTorch [here](https://pytorch.org/get-started/locally/) (for GPU users, **choose CUDA 11.0**) * Extra Dependencies: * Windows: * Install Visual C++ 2015 build tools [here](https://go.microsoft.com/fwlink/?LinkId=691126) and install default option. * Install latest Visual C++ redistributable [here](https://support.microsoft.com/en-us/help/2977003/the-latest-supported-visual-c-downloads) and choose x86 for 32 bit OS, x64 for 64 bit OS. * Linux: ``` bash $ apt-get install libglib2.0-0 libsm6 libxrender1 libxext6 ``` * Mac: ``` bash $ echo No extra dependency needed ":)" ``` ### 2. Install FastEstimator: * Stable (Linux/Mac): ``` bash $ pip install fastestimator ``` * Stable (Windows): First download zip file from [available releases](https://github.com/fastestimator/fastestimator/releases) ``` bash $ pip install fastestimator-x.x.x.zip ``` * Most Recent (Linux/Mac): ``` bash $ pip install fastestimator-nightly ``` * Most Recent (Windows): First download zip file [here](https://github.com/fastestimator/fastestimator/archive/master.zip) ``` bash $ pip install fastestimator-master.zip ``` ## Docker Hub Docker containers create isolated virtual environments that share resources with a host machine. Docker provides an easy way to set up a FastEstimator environment. You can simply pull our image from [Docker Hub](https://hub.docker.com/r/fastestimator/fastestimator/tags) and get started: * GPU: ``` bash docker pull fastestimator/fastestimator:latest-gpu ``` * CPU: ``` bash docker pull fastestimator/fastestimator:latest-cpu ``` ## Useful Links: * [Website](https://www.fastestimator.org): More info about FastEstimator API and news. * [Tutorial Series](https://github.com/fastestimator/fastestimator/tree/master/tutorial): Everything you need to know about FastEstimator. * [Application Hub](https://github.com/fastestimator/fastestimator/tree/master/apphub): End-to-end deep learning examples in FastEstimator. ## Citation Please cite FastEstimator in your publications if it helps your research: ``` @misc{fastestimator, title = {FastEstimator: A Deep Learning Library for Fast Prototyping and Productization}, author = {Xiaomeng Dong and Junpyo Hong and Hsi-Ming Chang and Michael Potter and Aritra Chowdhury and Purujit Bahl and Vivek Soni and Yun-Chan Tsai and Rajesh Tamada and Gaurav Kumar and Caroline Favart and V. Ratna Saripalli and Gopal Avinash}, note = {NeurIPS Systems for ML Workshop}, year = {2019}, url = {http://learningsys.org/neurips19/assets/papers/10_CameraReadySubmission_FastEstimator_final_camera.pdf} } ``` ## License [Apache License 2.0](https://github.com/fastestimator/fastestimator/blob/master/LICENSE)