[Feat] Setup GPU environments

[GangBean]

Background

• Need to setup GPU computing resources for reducing training and inference time.
• It's better to utilize free resources as possible.
• AWS Free tier offers you free GPU instance usage time.
• AWS SageMaker offers notebook instance which can be managed through AWS Console.
• However, it's not convenient to control versions through notebook files. So, concluded to setup environments using SageMaker Python SDK, with which you can use AWS GPU computing power without python notebook files.

ToDo

• [ ] make training environment with SageMaker using GPU instance

[GangBean]

• There is no GPU computing power instance among free tier instances

[GangBean]

• Need to find another free GPU resource.

[GangBean]

• GCP provides $300 free credits for 90 days.
• GCP g2 instances include nvidia L4 GPU.