Ideas

Showing 44 of 18234

mbatchd log job submission rejection

It is helpful to have job submission rejection logging to log file so we can set up monitor to size the issue and early warning alert of scale of issue

over 2 years ago in Spectrum LSF / Other 1 Not under consideration

1 VOTE

Provide ability to dump sequence of jobs ids considered by LSF scheduler for dispatch during a dispatch cycle

An old job which is pending in LSF may continue to remain pending with the reason "Individual host based reasons" while newer jobs continue to get scheduled. This may happen due to queue priority and fairshare settings. When running 'badmin diagno...

over 2 years ago in Spectrum LSF / Other 0 Future consideration

1 VOTE

Enabling the Batch processing based on Compute Express Link CXL 3.0 or CXL 4.0 Fabrics

CXL 3.0 is built upon the IBM OpenCAPI, but CXL3 and CXL 4 have more speeds and will achieve the following: 1- Enabling the Interconnection between IBM LinuxONE and Exa scale AI Data Centers based on Nvidia DGX H100 based on Nvidia H100 GPUs and I...

almost 3 years ago in Spectrum LSF / Other 1 Not under consideration

1 VOTE

Adding heterogeneous processing CPUs, GPUs and FPGA Accelerator for Spectrum LSF

In the near future we will need more advanced distributed heterogeneous processing , We think it is better to take into consideration the Intel oneAPI and other tools like oneDNN, oneDAL ,...etc. It may need joining openPOWER foundation and IBM wi...

almost 3 years ago in Spectrum LSF / Other 1 Not under consideration

2 VOTE

Add class labels to model metadata

Classification models can usually predict the probabilities per class. One probability number per known class. The model metadata in WML does not include the list of class labels seen in training. This makes it difficult for a scoring application ...

almost 6 years ago in Watson Machine Learning 1 Not under consideration

2 VOTE

Add Form POST support for web services

Taken from https://github.ibm.com/PrivateCloud/support/issues/716 attempt to call Python web service with file_upload() or any multi-part form POST call fails with web service error [call should invoke Python method] [customer is uploading ima...

almost 6 years ago in Watson Machine Learning 0 Not under consideration

1 VOTE

LSF should support output/error file path that exists only in a container

As of now LSF cannot handle a scenario where the output/error file path exists only in the container and does not exist on the underlying host where the container is running .For example:bsub -o <path_in_container_only> --app <my_bapp>...

almost 3 years ago in Spectrum LSF / Other 1 Not under consideration

2 VOTE

Allow users to check CUH of Machine Learning by themselves.

-Use a real-life scenario to explain the problem statement/pain points, Customers can't check how many CUH they are using for Machine Learning. -state your current workaround(s), state any proposed solution(s), Keep a memo how many hours they used...

over 7 years ago in Watson Machine Learning 0 Functionality already exists

2 VOTE

Provide a detailed listing of differences between CLI, Python Client and UI screens

CLI, Python Client and UI screens differ in terms of maturity and capabilitie. It’s frustrating to start working with one of the tools and find out that the tool doesn’t yet have the capabilities you require, and then you have to switch to another...

over 7 years ago in Watson Machine Learning 1 Not under consideration

2 VOTE

Add functionality to export tensorflow (and keras) models as tensorflow.lite and tensorflow.js to support running on devices and in the browser

Tensorflow.lite allows running models on devices rather than having to use server side apis. Similarly, it would be great if we could export tensorflow.js code. This is similar in concept to Core ML but Core ML only supports ios.

over 7 years ago in Watson Machine Learning 0 Not under consideration

Please enter your email address

FILTER BY CATEGORY