Integration with IBM Watson Machine Learning and Watson Studio Local
WML Accelerator can work together with IBM Watson® Studio Local and Watson Machine Learning.
WML Accelerator can be installed with IBM Watson Studio Local which provides you with the capabilities of both offerings as well as the use of Watson Studio experiments.
For installation instructions, see Installing WML Accelerator.
Supported deep learning frameworks
The following deep learning frameworks work with WML Accelerator, WML and Watson Studio.
| WML Accelerator Version | WML Version | Watson Studio Local Version | Frameworks | Python Version |
|---|---|---|---|---|
| 1.2.1 You must apply interim fix 526695. |
2.0 The scoring and deployment function in WML is not available in this integration. |
2.0 | TensorFlow 1.14, PyTorch 1.1, IBM Caffe 1.0 All frameworks are available for training only. |
2, 3 |
| 1.2.1 You must apply interim fix 526695. |
2.0.1 The scoring and deployment function in WML is not available in this integration. |
2.0.1 | TensorFlow 1.14, PyTorch 1.1, IBM Caffe 1.0 All frameworks are available for training only. |
2, 3 |
| 1.2.1 You must apply interim fix 526695. |
2.0.2 The scoring and deployment function in WML is not available in this integration. |
2.0.2 | TensorFlow 1.14, PyTorch 1.1, IBM Caffe 1.0 All frameworks are available for training only. |
2, 3 |
| 1.2.1 | IBM Cloud Pak for Data 3.0.1 | IBM Cloud Pak for Data 3.0.1 | TensorFlow 1.15, PyTorch 1.2, IBM Caffe 1.0 | 3.6 |
Resources and tutorials
Learn more about the Python library which allows WML Accelerator to work with the Watson Machine Learning on IBM Cloud, see http://wml-api-pyclient-dev-v4.mybluemix.net.
Learn more about using the WML API, see http://watson-ml-v4-api.mybluemix.net.