[NOT A BUG] : Is it possible to run the current validation/training scripts on a ray cluster?

[SaiKrishnaBala]

Python Version

python 3.11

Pip Freeze

I would be all the required packages to be installe din the runtime environment. Example: 
ray job submit     --address <address> --working-dir .      --runtime-env-json='{"pip": ["fire", "simple-parsing", "pyyaml", "mistral-common>=1.3.1", "safetensors", "tensorboard", "tqdm", "torch==2.2", "triton==2.2", "xformers==0.0.24"], "excludes": ["./mistral_models/"]}' -- python3 driver.py

Reproduction Steps

I have a ray cluster(A2) setup and I want to submit the training scripts as a job.

Expected Behavior

What changes are required to submit the current scripts to a ray cluster?

Additional Context

No response

Suggested Solutions

No response