Changes from #6 @chainyo

• Add StableDiffusionImg2ImgPipeline, StableDiffusionImageVariationPipeline, StableDiffusionUpscalePipeline along the StableDiffusionPipeline
• Add the task parameter to choose which pipeline to run
• Change pipeline output type from np.ndarray -> PIL.Image
• Add a download function to download the image when it's used as input
• Change image return format JPEG -> PNG
• Add an input pipeline check to ensure prompt/image is present for each pipeline
• Add a GET endpoint to check the API task status
• Add a way for the bot to check the API task loaded and raise an error if the task is not implemented for the bot