Torch-Pruning but with capabilities of supporting Brevitas QAT Modules

Does your model use Brevitas modules? And you want to prune it? What a coincidence! Use this repository functionalities to do just that!

Added capabilities: evaluating the reliability of deep learning models from hardware faults, tested on NVDLA machines

Name		Name	Last commit message	Last commit date
Latest commit History 1,502 Commits
.github/workflows		.github/workflows
FIdelity-Q		FIdelity-Q
FIdelityFramework		FIdelityFramework
assets		assets
examples		examples
models		models
reproduce		reproduce
tests		tests
torch_pruning		torch_pruning
.gitignore		.gitignore
AUTHORS		AUTHORS
LICENSE		LICENSE
README.md		README.md
accuracy.py		accuracy.py
finetune.py		finetune.py
get_layer_names.py		get_layer_names.py
lenet5_accuracy_vs_macs.png		lenet5_accuracy_vs_macs.png
lenet5_accuracy_vs_params.png		lenet5_accuracy_vs_params.png
lenet5_accuracy_vs_pruning.png		lenet5_accuracy_vs_pruning.png
lenetest.py		lenetest.py
plot_accuracy_graph.py		plot_accuracy_graph.py
prune.py		prune.py
prune_lenet5.py		prune_lenet5.py
requirements.txt		requirements.txt
resnet18_accuracy_vs_pruning.png		resnet18_accuracy_vs_pruning.png
resnet50_accuracy_vs_pruning.png		resnet50_accuracy_vs_pruning.png
setup.py		setup.py
test.py		test.py
test2.py		test2.py
test_accuracy.py		test_accuracy.py
test_pruned_loading.py		test_pruned_loading.py
thesis.docx		thesis.docx

Provide feedback