How to Deploy DeepSeek R1 Model with Amazon Bedrock | Step by Step guide

In this video, we'll walk you through the process of deploying your DeepSeek model on Amazon Bedrock, a cutting-edge AI platform on AWS.
Step 1: Install Git Large File Storage (LFS) - Done
curl -s packagecloud.i... | sudo bash
sudo apt install git-lfs
git lfs install
git lfs version
Step 2: Clone the DeepSeek Model Repository
git clone huggingface.co...
Step 3 :
aws s3 mb s3://bedrock-deepseek-models-llama-demo1-accountid-region
Step 4: Upload Model to S3
aws s3 sync --exclude '.git*' DeepSeek-R1-Distill-Llama-8B s3://bedrock-deepseek-models-llama-demo1--accountid-region/DeepSeek-R1-Distill-Llama-8B/

[@pspkutube]

Thanks! what would be the cost to run this on aws? I am looking at just playing it with for a day and then stopping the instance. I am assuming the cost of S3 and the some of the instance storage costs will be what I need to pay for till i delete it completely?
Also, what about an API like what openAI offers? there are some models that provide this for llama...

[@cloudncoffeewithnavnit]

Please review this blog - it covers the cost but your assumption is right - community.aws/content/2sIJqPaPMtmNxlRIQT5CzpTtziA/deploy-deepseek-r1-on-aws-bedrock

[@pspkutube]

@ Thank you

[@DaniyaMuzammil-x2t]

please provide us with complete code, some parts are not clear, for example from where are you copy pasting the sync command and make bucket, provide the syntax please

[@cloudncoffeewithnavnit]

Yes, Included in description of video

[@jigarsoni2319]

Hello I am facing ... ERROR
"The Safetensor files that you provided are not valid. Make sure the model weights are valid and that you can load them with the Safetensor method."
import is getting failed .. anyone can help mw with that ?

[@cloudncoffeewithnavnit]

I am wondering if which model you are looking to bring .. by looking at error messages it seems some issue with one of the file within git

[@cloudncoffeewithnavnit]

Could you please let me know which model you are trying to import ?

[@munkkis4n]

A bit too hectic. you jump between things so fast it is hard to follow.. You could have followed Chris Sandrini's blog post in order. It is clear and easy to follow but your video makes things hard to follow.

[@cloudncoffeewithnavnit]

Thanks for your feedback.. will definitely work it. Meanwhile feel free to use the Cris blog and instructions added in description.