Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Feature/sg 1215 update ptq and qat training #1618

Merged
merged 3 commits into from
Nov 7, 2023
Merged

Feature/sg 1215 update ptq and qat training #1618

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
952832e
Added PTQ and QAT for classification
BloodAxe Nov 7, 2023
ca77364
Update README
BloodAxe Nov 7, 2023
a09ec94
Merge branch 'master' into feature/SG-1215-update-ptq-and-qat-training
shaydeci Nov 7, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,7 @@ NOTEBOOKS_TO_RUN += src/super_gradients/examples/model_export/models_export_pose
NOTEBOOKS_TO_RUN += notebooks/what_are_recipes_and_how_to_use.ipynb
NOTEBOOKS_TO_RUN += notebooks/transfer_learning_classification.ipynb
NOTEBOOKS_TO_RUN += notebooks/how_to_use_knowledge_distillation_for_classification.ipynb
NOTEBOOKS_TO_RUN += notebooks/PTQ_and_QAT_for_classification.ipynb

# If there are additional notebooks that must not be executed, but still should be checked for version match, add them here
NOTEBOOKS_TO_CHECK := $(NOTEBOOKS_TO_RUN)
Expand Down
133 changes: 31 additions & 102 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -206,138 +206,67 @@ model = models.get("model-name", pretrained_weights="pretrained-model-name")
```
### Classification

#### Transfer Learning
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/transfer_learning_classification.ipynb"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Classification Transfer Learning</a>
</td>
</table>
</br></br>

* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/transfer_learning_classification.ipynb) [Transfer Learning for classification](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/transfer_learning_classification.ipynb)
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/PTQ_and_QAT_for_classification.ipynb) [PTQ and QAT for classification](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/PTQ_and_QAT_for_classification.ipynb)

### Semantic Segmentation

#### Quick Start
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3qKx9m8"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Segmentation Quick Start</a>
</td>
</table>
</br></br>



#### Transfer Learning
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3qKwMbe"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Segmentation Transfer Learning</a>
</td>
</table>
</br></br>



#### How to Connect Custom Dataset
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3QQBVJp"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Segmentation How to Connect Custom Dataset</a>
</td>
</table>
</br></br>

* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3qKx9m8) [Segmentation Quick Start](https://bit.ly/3qKx9m8)
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3qKwMbe) [Segmentation Transfer Learning](https://bit.ly/3qKwMbe)
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3QQBVJp) [How to Connect Custom Dataset](https://bit.ly/3QQBVJp)

### Pose Estimation

#### Transfer Learning

<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://colab.research.google.com/drive/1NMGzx8NdycIZqnRlZKJZrIOqyj0MFzJE#scrollTo=3UZJqTehg0On">
<img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Pose Estimation Transfer Learning
</a>
</td>
</table>
</br></br>
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1NMGzx8NdycIZqnRlZKJZrIOqyj0MFzJE#scrollTo=3UZJqTehg0On) [Pose Estimation Transfer Learning](https://colab.research.google.com/drive/1NMGzx8NdycIZqnRlZKJZrIOqyj0MFzJE#scrollTo=3UZJqTehg0On)


### Object Detection


#### Transfer Learning
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3SkMohx"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Detection Transfer Learning</a>
</td>
</table>
</br></br>

#### How to Connect Custom Dataset
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3dqDlg3"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Detection How to Connect Custom Dataset</a>
</td>
</table>
</br></br>


* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3SkMohx) [Object Detection Transfer Learning](https://bit.ly/3SkMohx)
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3dqDlg3) [How to Connect Custom Dataset](https://bit.ly/3SkMohx)

### How to Predict Using Pre-trained Model

#### Segmentation, Detection and Classification Prediction
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://bit.ly/3f4mssd"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> How to Predict Using Pre-trained Model</a>
</td>
</table>
</br></br>
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3f4mssd) [How to Predict Using Pre-trained Model](https://bit.ly/3f4mssd)



## Advanced Features
__________________________________________________________________________________________________________
### Post Training Quantization and Quantization Aware Training
Quantization involves representing weights and biases in lower precision, resulting in reduced memory and computational requirements, making it useful for deploying models on devices with limited resources. The process can be done during training, called Quantization aware training, or after training, called post-training quantization. A full tutorial can be found [here](http://bit.ly/41hC8uI).
<table class=“tfo-notebook-buttons” align=“left”>
<td width=“500”>
<a target="_blank" href="http://bit.ly/3KrN6an"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Post Training Quantization and Quantization Aware Training</a>
</td>
</table>
Quantization involves representing weights and biases in lower precision, resulting in reduced memory and computational requirements, making it useful for deploying models on devices with limited resources.
The process can be done during training, called Quantization aware training, or after training, called post-training quantization.
A full tutorial can be found [here](http://bit.ly/41hC8uI).

* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3KrN6an) [Post Training Quantization and Quantization Aware Training](https://bit.ly/3KrN6an)


### Quantization Aware Training YoloNAS on Custom Dataset
This tutorial provides a comprehensive guide on how to fine-tune a YoloNAS model using a custom dataset. It also demonstrates how to utilize SG's QAT (Quantization-Aware Training) support. Additionally, it offers step-by-step instructions on deploying the model and performing benchmarking.
<table class=“tfo-notebook-buttons” align=“left”>
<td width=“500”>
<a target="_blank" href="https://bit.ly/3MIKdTy"><img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Quantization Aware Training YoloNAS on Custom Dataset</a>
</td>
</table>
This tutorial provides a comprehensive guide on how to fine-tune a YoloNAS model using a custom dataset.
It also demonstrates how to utilize SG's QAT (Quantization-Aware Training) support. Additionally, it offers step-by-step instructions on deploying the model and performing benchmarking.

* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://bit.ly/3MIKdTy) [Quantization Aware Training YoloNAS on Custom Dataset](https://bit.ly/3MIKdTy)


### Knowledge Distillation Training

Knowledge Distillation is a training technique that uses a large model, teacher model, to improve the performance of a smaller model, the student model.
Learn more about SuperGradients knowledge distillation training with our pre-trained BEiT base teacher model and Resnet18 student model on CIFAR10 example notebook on Google Colab for an easy to use tutorial using free GPU hardware
<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/how_to_use_knowledge_distillation_for_classification.ipynb">
<img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> Knowledge Distillation Training
</a>
</td>
</table>
</br></br>

* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/how_to_use_knowledge_distillation_for_classification.ipynb) [Knowledge Distillation Training](https://github.com/Deci-AI/super-gradients/blob/master/notebooks/how_to_use_knowledge_distillation_for_classification.ipynb)


### Recipes
To train a model, it is necessary to configure 4 main components. These components are aggregated into a single "main" recipe `.yaml` file that inherits the aforementioned dataset, architecture, raining and checkpoint params. It is also possible (and recomended for flexibility) to override default settings with custom ones.

To train a model, it is necessary to configure 4 main components.
These components are aggregated into a single "main" recipe `.yaml` file that inherits the aforementioned dataset, architecture, raining and checkpoint params.
It is also possible (and recomended for flexibility) to override default settings with custom ones.
All recipes can be found [here](http://bit.ly/3gfLw07)
</br>
Recipes support out of the box every model, metric or loss that is implemented in SuperGradients, but you can easily extend this to any custom object that you need by "registering it". Check out [this](http://bit.ly/3TQ4iZB) tutorial for more information.

<table class="tfo-notebook-buttons" align="left">
<td width="500">
<a target="_blank" href="https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/what_are_recipes_and_how_to_use.ipynb">
<img src="https://github.com/Deci-AI/super-gradients/raw/master/documentation/assets/SG_img/colab_logo.png" /> How to Use Recipes
</a>
</td>
</table>
</br></br>

</br>
* [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/Deci-AI/super-gradients/blob/master/notebooks/what_are_recipes_and_how_to_use.ipynb) [How to Use Recipes](https://github.com/Deci-AI/super-gradients/blob/master/notebooks/what_are_recipes_and_how_to_use.ipynb)

<details markdown="1">
<summary><h3>Using Distributed Data Parallel (DDP) </h3></summary>

Expand Down
1,857 changes: 1,857 additions & 0 deletions notebooks/PTQ_and_QAT_for_classification.ipynb
View file
Open in desktop

Large diffs are not rendered by default.