Explanation:
The most straightforward and efficient method to install the gpu_ml library on a Databricks GPU-enabled cluster is by executing %pip install gpu_ml within a notebook attached to the cluster. This approach is preferred because:
gpu_ml to the cluster's GPU library dependencies is not required since the library can be directly installed via the notebook.PYTHON_GPU_LIB variable is not advised as it may cause conflicts with other GPU libraries.%pip command.
Thus, using the %pip install gpu_ml command within the notebook is the recommended approach for enabling GPU-accelerated computations.Ultimate access to all questions.
No comments yet.
A data scientist is looking to utilize a GPU-accelerated version of a machine learning library, gpu_ml, on a Databricks GPU-enabled cluster. What is the most effective method to install this library?
A
Modify the cluster to utilize the Databricks Runtime for GPU.
B
Include gpu_ml in the cluster's GPU library dependencies.
C
Execute %pip install gpu_ml in any notebook connected to the cluster.
D
Configure the PYTHON_GPU_LIB variable in the cluster settings to include gpu_ml.
E
GPU-accelerated libraries cannot be installed on a Databricks cluster.