Adapt an existing CLM Tuning script to work with GPU parallelism

Job ID: 35342302

Budget: $10 – $30 CAD

finetuning hugging face models with the CLM script but I need to implement GPU parallelism in order to get lower eval losses and train bigger models.
possibly with Deep speed accelerator.config but am open to other options
Related categories: Python GPGPU CUDA Machine Learning (ML) Deep Learning