- 최신
- 최다 투표
- 가장 많은 댓글
Hi,
Yes, Bedrock models can be trained on customer data with full privacy protected. See https://aws.amazon.com/blogs/aws/customize-models-in-amazon-bedrock-with-your-own-data-using-fine-tuning-and-continued-pre-training/
On training costs, the blog above says:
Model customization is charged per tokens processed. This is the number of tokens in the training
dataset multiplied by the number of training epochs. An epoch is one full pass through the training
data during customization. Model storage is charged per month, per model. Inference is charged hourly
per model unit using provisioned throughput. For detailed pricing information, see Amazon Bedrock Pricing.
But, currently Stable AI models are not supported for fine tuning: see https://aws.amazon.com/bedrock/pricing/
Best,
Didier
Yes you can finetune Stable Diffusion models on both Bedrock and Sagemaker :
Bedrock :
You may want to focus on the minute 32:50 For pricing you may see this :
What is Amazon Bedrock? - Amazon Bedrock pricing - https://docs.aws.amazon.com/bedrock/latest/userguide/what-is-bedrock.html#bedrock-pricing
SageMaker :
The Bedrock video mentioned in this reply describes training SD via Sagemaker Jumpstart - i don't believe there is a way now to fine tune and train SD directly in Bedrock
thank you, I am exploring AWS SageMaker, does it support training Stable AI models?
Hi, You're welcome. Thanks for accepting my answer. Amazon Bedrock supports some StableAI models: https://aws.amazon.com/bedrock/stable-diffusion/ But, AFAIK, you have to use them via Amazon Bedrock only. You can't launch your own instance of StableAI in Sagemaker