Skip to content

Model Zoo - TensorFlow

The RBLN TensorFlow model zoo provides a variety of neural network models designed to run on the RBLN NPU. The number of models supported by the RBLN model zoo will continuously expand as the RBLN SDK is updated. You can access full list of the models in the RBLN Model Zoo GitHub repository.

Supported models

The table shows the full list of the models covered by the RBLN TensorFlow model zoo as of today.

Model Dataset Task
BERT-base - BookCorpus & English Wikipedia
- SQuAD v2
BERT-large - BookCorpus & English Wikipedia
- SQuAD v2
DeepLabV3+_ResNet50 ILSVRC2012 Semantic Segmentation
Unet-DenseNet201 NYU Depth Dataset V2 Monocular Depth Estimation
ConvNeXtTiny ILSVRC2012 Image Classification
ConvNeXtSmall ILSVRC2012 Image Classification
ConvNeXtBase ILSVRC2012 Image Classification
ConvNeXtLarge ILSVRC2012 Image Classification
ConvNextXLarge ILSVRC2012 Image Classification
EfficientNetB0 ILSVRC2012 Image Classification
EfficientNetB1 ILSVRC2012 Image Classification
EfficientNetB2 ILSVRC2012 Image Classification
EfficientNetB3 ILSVRC2012 Image Classification
EfficientNetB4 ILSVRC2012 Image Classification
EfficientNetB5 ILSVRC2012 Image Classification
EfficientNetB6 ILSVRC2012 Image Classification
EfficientNetB7 ILSVRC2012 Image Classification
EfficientNetV2B0 ILSVRC2012 Image Classification
EfficientNetV2B1 ILSVRC2012 Image Classification
EfficientNetV2B2 ILSVRC2012 Image Classification
EfficientNetV2B3 ILSVRC2012 Image Classification
EfficientNet_V2_S ILSVRC2012 Image Classification
EfficientNet_V2_M ILSVRC2012 Image Classification
EfficientNet_V2_L ILSVRC2012 Image Classification
NasNetLarge ILSVRC2012 Image Classification
NasNetMobile ILSVRC2012 Image Classification
MobileNet ILSVRC2012 Image Classification
MobileNet_V2 ILSVRC2012 Image Classification
MobileNet_V3_Small ILSVRC2012 Image Classification
MobileNet_V3_Large ILSVRC2012 Image Classification
ResNet50 ILSVRC2012 Image Classification
ResNet101 ILSVRC2012 Image Classification
ResNet152 ILSVRC2012 Image Classification
ResNet50V2 ILSVRC2012 Image Classification
ResNet101V2 ILSVRC2012 Image Classification
ResNet152V2 ILSVRC2012 Image Classification
VGG16 ILSVRC2012 Image Classification
VGG19 ILSVRC2012 Image Classification
DenseNet121 ILSVRC2012 Image Classification
DenseNet169 ILSVRC2012 Image Classification
DenseNet201 ILSVRC2012 Image Classification
RegNet_X_200MF ILSVRC2012 Image Classification
RegNet_X_400MF ILSVRC2012 Image Classification
RegNet_X_600MF ILSVRC2012 Image Classification
RegNet_X_800MF ILSVRC2012 Image Classification
RegNet_X_1_6GF ILSVRC2012 Image Classification
RegNet_X_3_2GF ILSVRC2012 Image Classification
RegNet_X_4GF ILSVRC2012 Image Classification
RegNet_X_6_4GF ILSVRC2012 Image Classification
RegNet_X_8GF ILSVRC2012 Image Classification
RegNet_X_12GF ILSVRC2012 Image Classification
RegNet_X_16GF ILSVRC2012 Image Classification
RegNet_X_32GF ILSVRC2012 Image Classification
RegNet_Y_200MF ILSVRC2012 Image Classification
RegNet_Y_400MF ILSVRC2012 Image Classification
RegNet_Y_600MF ILSVRC2012 Image Classification
RegNet_Y_800MF ILSVRC2012 Image Classification
RegNet_Y_1_6GF ILSVRC2012 Image Classification
RegNet_Y_3_2GF ILSVRC2012 Image Classification
RegNet_Y_4GF ILSVRC2012 Image Classification
RegNet_Y_6_4GF ILSVRC2012 Image Classification
RegNet_Y_8GF ILSVRC2012 Image Classification
RegNet_Y_12GF ILSVRC2012 Image Classification
RegNet_Y_16GF ILSVRC2012 Image Classification
RegNet_Y_32GF ILSVRC2012 Image Classification
Inception_V3 ILSVRC2012 Image Classification
InceptionResNetV2 ILSVRC2012 Image Classification
Xception ILSVRC2012 Image Classification
ResNetRS50 ILSVRC2012 Image Classification
ResNetRS101 ILSVRC2012 Image Classification
ResNetRS152 ILSVRC2012 Image Classification
ResNetRS200 ILSVRC2012 Image Classification
ResNetRS270 ILSVRC2012 Image Classification
ResNetRS350 ILSVRC2012 Image Classification
ResNetRS420 ILSVRC2012 Image Classification

How to run

We summarized the commands for running the models in the RBLN TensorFlow model zoo.

Classification (TF Keras Applications)

The TF Keras Applications libary provides various pre-trained classification backbone models. Here is the run command for the TF Keras Applications model DenseNet121:

1
2
3
$ cd rbln_model_zoo/tensorflow/vision/classification/keras_applications
$ python3 compile.py --model_name DenseNet121    # Model compile
$ python3 inference.py --model_name DenseNet121  # Inference

You can try with any of the models listed below by replacing the --model_name field of the run command:

ResNet50, ResNet101, ResNet152,
ResNet50V2, ResNet101V2, ResNet152V2,
EfficientNetB0, EfficientNetB1, EfficientNetB2,
EfficientNetB3, EfficientNetB4, EfficientNetB5,
EfficientNetB6, EfficientNetB7,
MobileNet, MobileNetV2, MobileNetV3Small, MobileNetV3Large,
EfficientNetV2B0, EfficientNetV2B1,
EfficientNetV2B2, EfficientNetV2B3,
EfficientNetV2L, EfficientNetV2M, EfficientNetV2S,
ConvNeXtSmall, ConvNeXtTiny, ConvNeXtBase, ConvNeXtLarge, ConvNeXtXLarge
RegNetX002, RegNetX004, RegNetX006, RegNetX008, RegNetX016, RegNetX032,
RegNetX040, RegNetX064, RegNetX080, RegNetX120, RegNetX160, RegNetX320,
RegNetY002, RegNetY004, RegNetY006, RegNetY008, RegNetY016, RegNetY032,
RegNetY040, RegNetY064, RegNetY080, RegNetY120, RegNetY160, RegNetY320,
DenseNet121, DenseNet169, DenseNet201,
NASNetLarge, NASNetMobile, InceptionV3,
VGG16, VGG19, InceptionResNetV2, Xception
ResNetRS101, ResNetRS152, ResNetRS200,
ResNetRS270, ResNetRS350, ResNetRS420, ResNetRS50

Language (Transformer)

Transformers have been widely used in natural language processing tasks such as laguage modeling, question-answering, and translation. You can run BERT, one of the most popular transformer models, with the RBLN TensorFlow model zoo.

You will see mlm and qa directories in the rbln_model_zoo/tensorflow/nlp/bert. mlm and qa respectively stand for 'masked language modeling' and 'question answering' tasks. Move to the directory that you want to test and run the following command with the model name:

$ python3 compile.py --model_name MODEL_NAME      # Model compilation
$ python3 inference.py --model_name MODEL_NAME    # Inference

The MODEL_NAME field in the above command can be one of the following list:

 bert_base, bert_large

Segmentation

You can run DeepLabV3+ for the multiclass semantic segmentation task with the following commands:

1
2
3
$ cd rbln_model_zoo/tensorflow/vision/segmentation
$ python3 compile.py     # Model compilation
$ python3 inference.py   # Inference

Depth Estimation

Here is the run command for the monocular depth estimation model UNet-DensNet201:

1
2
3
$ cd rbln_model_zoo/tensorflow/vision/depth_estimation
$ python3 compile.py     # Model compilation
$ python3 inference.py   # Inference