Build Low Code Automated Tensorflow, What-IF explainable models in just 3 lines of code. To make Deep Learning on Tensorflow absolutely easy for the masses with its low code framework and also increase trust on ML models through What-IF model explainability.
Project description
Auto Tensorflow Mission:
Build Low Code Automated Tensorflow, What-IF explainable models in just 3 lines of code.
To make Deep Learning on Tensorflow absolutely easy for the masses with its low code framework and also increase trust on ML models through What-IF model explainability.
Under the hood:
Built on top of the powerful Tensorflow ecosystem tools like TFX , TF APIs and What-IF Tool , the library automatically does all the heavy lifting internally like EDA, schema discovery, feature engineering, HPT, model search etc. This empowers developers to focus only on building end user applications quickly without any knowledge of Tensorflow, ML or debugging. There is no dependency on Pandas / SKLearn or other libraries which makes the whole pipeline highly scalable on any volume of data. Moreover the models trained with auto-tensorflow can directly be deployed on any cloud like GCP / AWS / Azure.
Official Launch: https://youtu.be/sil-RbuckG0
Features:
- Build Classification / Regression models on CSV data
- Automated Schema Inference
- Automated EDA and visualization for automated feature engineering
- Automated Model build for mixed data types( Continuous, Categorical and Free Text )
- Automated Hyper-parameter tuning
- Automated GPU Distributed training
- Automated UI based What-IF analysis
- Control over complexity of model
- No dependency over Pandas / SKLearn
- Can handle dataset of any size - including multiple CSV files
Tutorials:
Setup:
- Install library using -
pip install auto-tensorflow
orpip install git+https://github.com/rafiqhasan/auto-tensorflow.git
- Works best on UNIX/Linux/Debian/Google Colab/MacOS
Usage:
- Initialize TFAuto Engine
from auto_tensorflow.tfa import TFAuto
tfa = TFAuto(train_data_path='/content/train_data/', test_data_path='/content/test_data/', path_root='/content/tfauto')
- Step 1 - Automated EDA and Schema discovery
tfa.step_data_explore(viz=True) ##Viz=False for no visualization
- Step 2 - Automated ML model build and train
tfa.step_model_build(label_column = 'price', model_type='REGRESSION', model_complexity=1)
- Step 3 - Automated What-IF Tool launch
tfa.step_model_whatif()
API Arguments:
-
Method TFAuto
train_data_path
: Path where training data is storedtest_data_path
: Path where Test / Eval data is storedpath_root
: Directory for running TFAuto( Directory should NOT exist )
-
Method step_data_explore
viz
: Is data visualization required ? - True or False( Default )
-
Method step_model_build
label_column
: The feature to be used as Labelmodel_type
: Either of 'REGRESSION'( Default ), 'CLASSIFICATION'model_complexity
:0
: Model with default hyper-parameters1
(Default): Model with automated hyper-parameter tuning2
: Complexity 1 + Automated tuning of Text layers
Current limitations:
There are a few limitations in the initial release but we are working day and night to resolve these and add them as future features.
- Doesn't support Image / Audio data
- Doesn't support - quote delimited CSVs( TFX doesn't support qCSV yet )
Future roadmap:
- Add support for Timeseries / Audio / Image data
- Add feature to download full pipeline model Python code for advanced tweaking
Release History:
1.1.1 - 14/07/2021
- Fixed bugs
- Added more features:
- Added complexity = 2 for automated tunable textual layers
- Textual label for Classification
- Imbalanced label handling
- GPU fixes
1.0.1 - 07/07/2021
- First public release
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for auto_tensorflow-1.1.1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 05af6889f8aaa2e4cce534bc10d5b923f17791d440162080e6dc742958c09178 |
|
MD5 | 3be69e277ada50404f80d8f590d6d49f |
|
BLAKE2b-256 | c637f2d0e6117456b04ee9d0992455536191dc8268d411a677935754efa1d894 |