Supported deep learning frameworks
Supported deep learning frameworks included with IBM Watson® Machine Learning Accelerator.
The following frameworks are included with IBM Watson Machine Learning Accelerator for use with IBM Watson Machine Learning.
Framework, library or asset | Version available in Refresh 9 | Version available in Refresh 8 | Version available in Refresh 7 | Version available in Refresh 6 | Version available in Refresh 5 | Version available in Refresh 4 | Version available in Refresh 3 | Version available in Refresh 2 | Version available in Refresh 1 | Version available in WML Accelerator 2.3 |
---|---|---|---|---|---|---|---|---|---|---|
TensorFlow | 2.7.1 | 2.7.1 | 2.7.0 | 2.7.0 | 2.4.4 | 2.4.4 | 2.4.3 | 2.4.3 | 2.4.2 | 2.4.1 |
PyTorch | 1.10.2 | 1.10.2 | 1.10.1 | 1.10.1 | 1.7.1 | 1.7.1 | 1.7.1 | 1.7.1 | 1.7.1 | 1.7.1 |
Tensorboard | 2.7.0 | 2.7.0 | 2.7.0 | 2.7.0 | 2.4.1 | 2.4.1 | 2.4.1 | 2.4.1 | 2.4.1 | 2.4.1 |
TensorRT | Not applicable | Not applicable | Not applicable | Not applicable | Not applicable | Not applicable | Not applicable | Not applicable | Not applicable | 7.2.1.6 |
torchVision | 0.11.3 | 0.11.3 | 0.11.2 | 0.11.2 | 0.8.2 | 0.8.2 | 0.8.2 | 0.8.2 | 0.8.2 | 0.8.2 |
OpenCV | 4.5.5 | 4.5.5 | 4.5.3 | 4.5.3 | 3.4.10 | 3.4.10 | 3.4.10 | 3.4.10 | 3.4.10 | 3.4.10 |
scikit-learn | 1.0.2 | 1.0.2 | 1.0.2 | 1.0.2 | 0.24.2 | 0.24.2 | 0.24.2 | 0.24.2 | 0.24.2 | 0.23.2 |
XGBoost | 1.5.2 | 1.5.2 | 1.5.1 | 1.5.1 | 1.3.3 | 1.3.3 | 1.3.3 | 1.3.3 | 1.3.3 | 1.3.3 |
ONNX | 1.10.2 (tf2onnx 1.9.3, skl2onnx 1.10.4) | 1.10.2 (tf2onnx 1.9.3, skl2onnx 1.10.4) | 1.10.2 (tf2onnx 1.9.3, skl2onnx 1.10.3) | 1.10.2 (tf2onnx 1.9.3, skl2onnx 1.10.3) | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 | 1.6.0 |
PyArrow | 5.0.0 | 5.0.0 | 5.0.0 | 5.0.0 | 3.0.0 | 3.0.0 | 3.0.0 | 3.0.0 | 3.0.0 | 3.0.0 |
Python | 3.9.7 | 3.9.7 | 3.9.7 | 3.9.7 | 3.8.12 | 3.8.12 | 3.8.11 | 3.8.11 | 3.8.10 | 3.7.10 |
- All frameworks support the supported versions of Python.
- PyTorch supports single node training, distributed training, and elastic distributed training.
- TensorFlow supports single node training, distributed training, and elastic distributed training (with TF.Keras).