Skip to content

Commit ea32d89

Browse files
atierianatierian
committed
ai kit: add table for supported models and add nova
1 parent a357473 commit ea32d89

File tree

1 file changed

+126
-45
lines changed
  • src/pages/[platform]/ai/concepts/models

1 file changed

+126
-45
lines changed

‎src/pages/[platform]/ai/concepts/models/index.mdx

Lines changed: 126 additions & 45 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,5 @@
11
import { getCustomStaticPath } from "@/utils/getCustomStaticPath";
2+
import { Table, TableBody, TableCell, TableHead, TableRow } from '@aws-amplify/ui-react';
23

34
export const meta = {
45
title: "Models",
@@ -30,12 +31,10 @@ export function getStaticProps(context) {
3031

3132
A foundation model is a large, general-purpose machine learning model that has been pre-trained on a vast amount of data. These models are trained in an unsupervised or self-supervised manner, meaning they learn patterns and representations from the unlabeled training data without being given specific instructions or labels.
3233

33-
Foundation models are useful because they are general-purpose and you don't need to train the models yourself, but are powerful enough to take on a range of applications.
34+
Foundation models are useful because they are general-purpose and you don't need to train the models yourself, but are powerful enough to take on a range of applications.
3435

3536
Foundation Models, which Large Language Models are a part of, are inherently stateless. They take input in the form of text or images and generate text or images. They are also inherently non-deterministic. Providing the same input can generate different output.
3637

37-
38-
3938
## Getting model access
4039

4140
Before you can invoke a foundation model on Bedrock you will need to [request access to the models in the AWS console](https://console.aws.amazon.com/bedrock/home#/modelaccess).
@@ -44,51 +43,135 @@ Be sure to check the region you are building your Amplify app in!
4443

4544
## Pricing and Limits
4645

47-
Each foundation model in Amazon Bedrock has its own pricing and throughput limits for on-demand use. On-demand use is serverless, you don't need to provision any AWS resources to use and you only pay for what you use. The Amplify AI kit uses on-demand use for Bedrock.
46+
Each foundation model in Amazon Bedrock has its own pricing and throughput limits for on-demand use. On-demand use is serverless, you don't need to provision any AWS resources to use and you only pay for what you use. The Amplify AI kit uses on-demand use for Bedrock.
4847

49-
The cost for using foundation models is calculated by token usage. A token in generative AI refers to chunks of data that were sent as input and how much data was generated. A token is roughly equal to a word, but depends on the model being used. Each foundation model in Bedrock has its own pricing based on input and output tokens used.
48+
The cost for using foundation models is calculated by token usage. A token in generative AI refers to chunks of data that were sent as input and how much data was generated. A token is roughly equal to a word, but depends on the model being used. Each foundation model in Bedrock has its own pricing based on input and output tokens used.
5049

5150
When you use the Amplify AI Kit, inference requests are charged to your AWS account based on Bedrock pricing. There is no Amplify markup, you are just using AWS resources in your own account.
5251

5352
Always refer to [Bedrock pricing](https://aws.amazon.com/bedrock/pricing/) for the most up-to-date information on running generative AI with Amplify AI Kit.
5453

54+
Your Amplify project must be deployed to a region where the foundation model you specify is available. See [Bedrock model support](https://docs.aws.amazon.com/bedrock/latest/userguide/models-supported.html) for the supported regions per model.
5555

5656
## Supported Providers and Models
5757

58-
The Amplify AI Kit uses Bedrock's [Converse API](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html) to leverage a unified API across models. Most models have different structures to how they best work with input and how they format their output. For example, ...
59-
60-
### AI21 Labs
61-
* Jamba 1.5 Large
62-
* Jamba 1.5 Mini
63-
[Bedrock documentation about AI21 models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-ai21.html)
64-
65-
### Anthropic
66-
* Claude 3 Haiku
67-
* Claude 3.5 Haiku
68-
* Claude 3 Sonnet
69-
* Claude 3 Opus
70-
* Claude 3.5 Sonnet
71-
* Claude 3.5 Sonnet v2
72-
[Bedrock documentation about Anthropic models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html)
73-
74-
### Cohere
75-
* Command R
76-
* Command R+
77-
[Bedrock documentation about Cohere models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-cohere.html)
78-
79-
### Meta Llama
80-
* Llama 3.1
81-
[Bedrock documentation about Meta Llama models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-meta.html)
82-
83-
### Mistral AI
84-
* Large
85-
* Large 2
86-
[Bedrock documentation about Mistral AI models](https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-mistral.html)
87-
88-
89-
The Amplify AI Kit makes use of ["tools"](/[platform]/ai/concepts/tools) for both generation and conversation routes. [The models it supports must support tool use in the Converse API](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html).
90-
91-
Using the Converse API makes it easy to swap different models without having to drastically change how you interact with them.
58+
The Amplify AI Kit uses Bedrock's [Converse API](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html) to leverage a unified API across models.
59+
60+
<Table
61+
caption="Table with supported models for Amplify AI kit"
62+
highlightOnHover={false}
63+
style={{ border: '1.5px solid' }}>
64+
<TableHead>
65+
<TableRow>
66+
<TableCell as="th">Provider</TableCell>
67+
<TableCell as="th">Model</TableCell>
68+
<TableCell as="th">Conversation</TableCell>
69+
<TableCell as="th">Generation</TableCell>
70+
</TableRow>
71+
</TableHead>
72+
<TableBody>
73+
<TableRow>
74+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-ai21.html">AI21 Labs</a></strong></TableCell>
75+
<TableCell>Jurassic-2 Large</TableCell>
76+
<TableCell>✅</TableCell>
77+
<TableCell>❌</TableCell>
78+
</TableRow>
79+
<TableRow>
80+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-ai21.html">AI21 Labs</a></strong></TableCell>
81+
<TableCell>Jurassic-2 Mini</TableCell>
82+
<TableCell>✅</TableCell>
83+
<TableCell>❌</TableCell>
84+
</TableRow>
85+
<TableRow style={{ borderTop: '1.5px solid' }}>
86+
<TableCell><strong><a href="https://aws.amazon.com/ai/generative-ai/nova/">Amazon</a></strong></TableCell>
87+
<TableCell>Amazon Nova Pro</TableCell>
88+
<TableCell>✅</TableCell>
89+
<TableCell>❌</TableCell>
90+
</TableRow>
91+
<TableRow>
92+
<TableCell><strong><a href="https://aws.amazon.com/ai/generative-ai/nova/">Amazon</a></strong></TableCell>
93+
<TableCell>Amazon Nova Lite</TableCell>
94+
<TableCell>✅</TableCell>
95+
<TableCell>❌</TableCell>
96+
</TableRow>
97+
<TableRow>
98+
<TableCell><strong><a href="https://aws.amazon.com/ai/generative-ai/nova/">Amazon</a></strong></TableCell>
99+
<TableCell>Amazon Nova Micro</TableCell>
100+
<TableCell>✅</TableCell>
101+
<TableCell>❌</TableCell>
102+
</TableRow>
103+
<TableRow style={{ borderTop: '1.5px solid' }}>
104+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
105+
<TableCell>Claude 3 Haiku</TableCell>
106+
<TableCell>✅</TableCell>
107+
<TableCell>✅</TableCell>
108+
</TableRow>
109+
<TableRow>
110+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
111+
<TableCell>Claude 3.5 Haiku</TableCell>
112+
<TableCell>✅</TableCell>
113+
<TableCell>✅</TableCell>
114+
</TableRow>
115+
<TableRow>
116+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
117+
<TableCell>Claude 3 Sonnet</TableCell>
118+
<TableCell>✅</TableCell>
119+
<TableCell>✅</TableCell>
120+
</TableRow>
121+
<TableRow>
122+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
123+
<TableCell>Claude 3.5 Sonnet</TableCell>
124+
<TableCell>✅</TableCell>
125+
<TableCell>✅</TableCell>
126+
</TableRow>
127+
<TableRow>
128+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
129+
<TableCell>Claude 3.5 Sonnet v2</TableCell>
130+
<TableCell>✅</TableCell>
131+
<TableCell>✅</TableCell>
132+
</TableRow>
133+
<TableRow>
134+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-claude.html">Anthropic</a></strong></TableCell>
135+
<TableCell>Claude 3 Opus</TableCell>
136+
<TableCell>✅</TableCell>
137+
<TableCell>✅</TableCell>
138+
</TableRow>
139+
<TableRow style={{ borderTop: '1.5px solid' }}>
140+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-cohere.html">Cohere</a></strong></TableCell>
141+
<TableCell>Command R</TableCell>
142+
<TableCell>✅</TableCell>
143+
<TableCell>❌</TableCell>
144+
</TableRow>
145+
<TableRow>
146+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-cohere.html">Cohere</a></strong></TableCell>
147+
<TableCell>Command R+</TableCell>
148+
<TableCell>✅</TableCell>
149+
<TableCell>❌</TableCell>
150+
</TableRow>
151+
<TableRow style={{ borderTop: '1.5px solid' }}>
152+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-meta.html">Meta</a></strong></TableCell>
153+
<TableCell>Llama 3.1</TableCell>
154+
<TableCell>✅</TableCell>
155+
<TableCell>❌</TableCell>
156+
</TableRow>
157+
<TableRow style={{ borderTop: '1.5px solid' }}>
158+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-mistral.html">Mistral AI</a></strong></TableCell>
159+
<TableCell>Large</TableCell>
160+
<TableCell>✅</TableCell>
161+
<TableCell>❌</TableCell>
162+
</TableRow>
163+
<TableRow>
164+
<TableCell><strong><a href="https://docs.aws.amazon.com/bedrock/latest/userguide/model-parameters-mistral.html">Mistral AI</a></strong></TableCell>
165+
<TableCell>Large 2</TableCell>
166+
<TableCell>✅</TableCell>
167+
<TableCell>❌</TableCell>
168+
</TableRow>
169+
</TableBody>
170+
</Table>
171+
172+
Amplify AI Kit makes use of ["tools"](/[platform]/ai/concepts/tools) for both generation and conversation routes. [The models used must support tool use in the Converse API](https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference-supported-models-features.html).
173+
174+
Most models have different structures to how they best work with input and how they format their output. Using the Converse API makes it easy to swap different models without having to drastically change how you interact with them.
92175

93176
## Choosing a model
94177

@@ -100,17 +183,17 @@ Each model has its own context window size. The context window is how much infor
100183

101184
### Latency
102185

103-
Smaller models tend to have a lower latency than larger models, but can also sometimes be less powerful.
186+
Smaller models tend to have a lower latency than larger models, but can also sometimes be less powerful.
104187

105188
### Cost
106189

107-
Each model has its own price and throughput.
190+
Each model has its own price and throughput.
108191

109192
### Use-case fit
110193

111-
Some models are trained to be better at certain tasks or with certain languages.
194+
Some models are trained to be better at certain tasks or with certain languages.
112195

113-
Choosing the right model for your use case is balancing latency, cost, and performance.
196+
Choosing the right model for your use case is balancing latency, cost, and performance.
114197

115198

116199
## Using different models
@@ -136,5 +219,3 @@ const schema = a.schema({
136219
})
137220
})
138221
```
139-
140-

0 commit comments

Comments
 (0)