AWS Bedrock Generative AI: Train your own LLM with your own data

preview_player
Показать описание
Link to the full course
Рекомендации по теме
Комментарии
Автор

Hey there! Great explanation but I was still a bit confused on this.

So I am looking to use Bedrock to train a model on some internal company data so that we can ask questions and converse with it. But I am a bit confused still on the pricing.

So from what I see you have a few pricing things here:

The price to train your base model onto your custom data.
The price to store your custom model over time.
The price to actually access and use your custom model.

Could you help me with clarity on the last part?
So if you have a custom model, there is no way to access it on Demand. The way that it prices you is through Provisioned Throughput instead of "per x Tokens"?
I am very confused on that part.

Thank you though for the great video!

TheGreatJ
Автор

Can I use this to train a model on my own dataset (questions and answers) and answer a pdf in natural language?

sysadmin
Автор

Is the pricing per hour based on when a user sends a query? So every hour it’s x amount of dollars??

sysadmin
Автор

Hey, I am trying to build a chat bot with internal company data to ask questions, can i know how to store the history? and can i proceed in the same way you explained? my company wants to give commands or ask questions from bash

bharathkalyan