Live data from Hacker News

Amazon Nova

aws.amazon.com

81–90 of 157 posts

Re: Amazon Nova

#81
post #37

Earlier quoted context omitted.

You say they "get away with it," but it makes more sense to conclude that UI design has a lot lower ROI than we assume it does as users.

Oh I'm sure, the ACM UI was impossible to use for years to find certificates, they improved it, but, it will never have the same level of functionality that the API gives you and that's the bread and butter.

Imagine a native desktop app that let you build a UI with very basic elements, à la Visual Basic, and behind each of those elements is an associated AWS CLI command. Such that "aws s3 ls" attached to a list element would render an account's buckets.

The AWS APIs are so expansive, a product like this could offer a complete replacement for the default web console and maybe even charge for it. Does anyone know if such a solution exists? Perhaps some more generic "shell-to-ui" application? If not, I'm interested in building one if anybody would like to contribute.

Re: Amazon Nova

#82
post #79

Earlier quoted context omitted.

This doesn't include all the benchmarks. The one that really stands out is GroundUI-1K, where it beats the competition by 46%. Nova Pro looks like it could be a SOTA-comparable model at a lower price point.

SOTA?

"State of the Art", if that's what you were asking.

Re: Amazon Nova

#83

Earlier quoted context omitted.

TL;DR comparison of models vs frontier models on public benchmarks here https://imgur.com/a/CKMIhmm

This doesn't include all the benchmarks. The one that really stands out is GroundUI-1K, where it beats the competition by 46%. Nova Pro looks like it could be a SOTA-comparable model at a lower price point.

Just means it's better at one specific task than the others, which has always been the case. For each of Sonnet, GPT and Gemini I can readily name a task they are individually the best at. At the same time the consensus that Sonnet 3.5 is overall the currently strongest model remains correct, and that's what most people care about. Additionally most people do tasks that all of the models perform similarly at, or they can't be bothered to optimize every task by using the best model for that one task. Which makes sense since not a single cloud provider has all three of them. Now this one will likely be AWS-exclusive too.

Re: Amazon Nova

#84
post #71
post #67

Earlier quoted context omitted.

That may be generally true, but the linked page says Nova is a series of foundation models in the first sentence.

Yeah but even then they won't describe it using the same sort of language that everyone else developing these things does. How many parameters? What kind of corpus was it trained on? MoE, single model, or something else? Will the weights be available? It doesn't even use the words "LLM", "multimodal" or "transformer" which are clearly the most relevant terms here... "foundation model" isn't wrong but it's also the mo…

> How many parameters? What kind of corpus was it trained on?

It's rare for the leading model providers to answer these questions.

As someone who applies these models daily, I agree with the dead comment from meta_x_ai. Your questions are interesting/relevant to a person developing these models, but less important to the average person utilizing these models through Bedrock.

Re: Amazon Nova

#86

A rough idea of the price differences... Per 1k tokens Input | Output Amazon Nova Micro: $0.000035 | $0.00014 Amazon Nova Lite: $0.00006 | $0.00024 Amazon Nova Pro: $0.0008 | $0.0032 Claude 3.5 Sonnet: $0.003 | $0.015 Claude 3.5 Haiku: $0.0008 | $0.0004 Claude 3 Opus: $0.015 | $0.075 Source: AWS Bedrock Pricing https://aws.amazon.com/bedrock/pricing/

You have added another zero for Haiku, its output cost is $0.004

You're absolutely right, apologies!

Re: Amazon Nova

#87

It would be nice if this was a truly open source model like OLMo: https://venturebeat.com/ai/truly-open-source-llm-from-ai2-to...

Is it narrowly open source, or somewhat open source, in some way? Thanks for that link, anyway!

As far as I can tell Amazon’s nova is fully closed source. Maybe because their goal is to get you to pay them for hosting.

Re: Amazon Nova

#88

Earlier quoted context omitted.

TL;DR comparison of models vs frontier models on public benchmarks here https://imgur.com/a/CKMIhmm

This doesn't include all the benchmarks. The one that really stands out is GroundUI-1K, where it beats the competition by 46%. Nova Pro looks like it could be a SOTA-comparable model at a lower price point.

in the berkeley function calling it is similar than 4-o for multi turn while being way faster

Re: Amazon Nova

#89

It's not clear what the use cases are for this, who is it aimed at.

It seems to be that it is faster and cheaper with slightly lower than SOTA quality. There’s an emerging subset of AI companies building features on “good enough” lightweight models: https://www.wired.com/story/how-do-you-get-to-artificial-gen...

So I guess that’s who it’s for.

I’ve only spent an hour with it though obviously.

Re: Amazon Nova

#90
post #74

Setting up AWS so you can try it via Amazon Bedrock API is a hassle, so I made a step-by-step guide: https://ndurner.github.io/amazon-nova . It's 14+ steps!

Thanks for that. Are there any proxies that can communicate with bedrock and serve it via a OpenAI style api?
Post reply on HN