[aws] setup finetuning pipeline for text-to-image model

[mytnguyen26]

We need a finetuning pipeline for text-to-image model

[mytnguyen26]

Refs:

• https://arxiv.org/abs/2112.10752
• https://huggingface.co/blog/stable_diffusion#how-does-stable-diffusion-work
• https://arxiv.org/pdf/2011.05552.pdf
• https://huggingface.co/docs/transformers/v4.39.0/en/model_doc/chinese_clip#chinese-clip
• https://huggingface.co/docs/accelerate/index

Code and training Ref:

• https://github.com/huggingface/diffusers/blob/main/examples/text_to_image/train_text_to_image.py#L144
• https://huggingface.co/docs/diffusers/v0.27.2/en/training/text2image