Generally, there is a small accuracy loss after quantization, but for some networks such
as Mobilenets, the accuracy loss can be large. In this situation, quantize finetuning
can be used to further improve the accuracy of quantized models.
Finetuning is almost the same as model training, which needs the original training dataset and a solver.prototxt. Follow the steps below to start finetuning with the fix_train_test.prototxt and caffemodel.
- Assign the training dataset to the input layer of fix_train_test.prototxt.
- Create a solver.prototxt file for
finetuning. An example of a solver.prototxt
file is provided below. You can adjust the hyper-parameters to get good results.
The most important parameter is base_lr, which is usually much smaller than the
one used in training.
- Run the following command to start
finetuning:
./vai_q_caffe finetune -solver solver.prototxt -weights quantize_results/quantize_train_test.caffemodel -gpu all - Deploy the finetuned model. The finetuned model is generated in the snapshot_prefix settings of the solver.prototxt file, such as ${snapshot_prefix}/finetuned_iter10000.caffemodel. You can use the test command to test its accuracy.
- Finally, you can use the deploy command to generate the deploy model (prototxt
and caffemodel) for the Vitis AI
compiler.
./vai_q_caffe deploy -model quantize_results/quantize_train_test.prototxt -weights finetuned_iter10000.caffemodel -gpu 0 -output_dir deploy_output