Using pre-trained model for Deep Learning and Machine Learning projects and application is known as Transfer Learning.
The models that have been introduced below has been developed by Machine Learning And Deep Learning Experts for many years and have been performed as the state-of-the-art
All the files that have been used here have been taken (and modified) from a free course on Deep Learning offered by Udacity (You can check it here - https://www.udacity.com/course/intro-to-tensorflow-for-deep-learning--ud187)
In this, You are going to learn how to load and use the state-of-the-art pre-trained models for your deep learning and machine learning projects or applications. The models that you can use for your DL/ML projects or applications have been loaded from Tensorflow Pre-Trained Model hub. You can view them here - https://tfhub.dev/ The models that are included on Tensorflow Hub are :
Refer to the given link for more details - https://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent
I. For Image Classification -
- EfficientNet
- MobileNet
- ImageNet
- ResNet
- Faster_RCNn You can view all the models here - https://tfhub.dev/s?module-type=image-augmentation,image-classification,image-feature-vector,image-generator,image-object-detection,image-others,image-style-transfer,image-rnn-agent
II. Text Classification -
- NNLM
- BERT
- ELMO
- LABSE You can view all the models here : https://tfhub.dev/s?module-type=text-embedding
In this colab,
- You will be introduced to Tensorflow Hub.
- You are going to use a state-of-the-art pre-trained model "MobileNet" for Flower-Set Dataset Classification task.
- This will a simple introduction to the Transfer Learning.
In this Colab,
- You will find yourself diving in the power of MobileNet
- You are going to use the model for the infamous Cats/Dogs Dataset Classification task.
In this Colab,
- You will work with Beans dataset taken from the tensorflow hub
Feel Free to Modify the repository.