Page History

...

var rp = ResourceProfile.upsert({'id': 'YourProfileName', 'name': 'YourProfileName', 'cpuCount': <cpus>, 'memoryMb': <cpu_memory>, 'diskGb': <disk_memory>, 'gpuCount': <gpus>, 'gpuMemoryMb': <gpu_memory>});

In addition to that, to effectively use the total memory requested in your resource profile, you will need to change the memory limits that are set at the user level of your Jupyter type. This can be done from the static console with:

...

Keep in mind that, currently, the maximum number of GPUs that can be used on the training cluster is one. If you request more than one, your static console is going to be unresponsive for some time, and you won't be able to start your Jupyter service. The same thing will happen if you request resources that surpass the limits of the current underlying hardware infrastructure or the limits set by C3 Admin. These are summarized in the table below:

...

If you are not sure about how to make your choice, but you know you need more than 1 CPU core, we recommend using the QuarterNode_cpu Resource Profile:

var rp = ResourceProfile.upsert({'id': 'QuarterNode_cpu', 'name': 'QuarterNode_cpu', 'cpuCount': 4, 'memoryMb': 16*1024, 'diskGb': 512, 'gpuCount': 0, 'gpuMemoryMb': 0});

Jupyter.setMemoryOverride(16*1024, ConfigOverride.USER);

Jupyter.startWithConfig({'resourceProfile': 'QuarterNode_cpu'});

If you are using GPUs, we recommend using the entire node, since each node has only one GPU. You can do that by creating the FullNode_gpu Resource Profile:

var rp = ResourceProfile.upsert({'id': 'FullNode_gpu', 'name': 'FullNode_gpu', 'cpuCount': 5, 'memoryMb': 48*1024, 'diskGb': 512, 'gpuCount': 1, 'gpuMemoryMb': 12*1024});

Jupyter.setMemoryOverride(12*1024, ConfigOverride.USER);

Jupyter.startWithConfig({'resourceProfile': 'FullNode_gpu'});

If you have additional requirements, please contact DTI DevOps.

Access Through IDS

Through IDS, clear controls regarding Jupyter are exposed under the 'ML Studio' section for Application. When you've selected a project, you need to start the Jupyter service to access notebooks on that project. In the creation process, you're able to select the container profile to use for your Jupyter service. By default, the 'BasicGpu' profile offers a single K80 GPU.

...