Comments (5)
Nice idea. I'm working on implementing a new BedrockConverseApi
instead of the current AbstractBedrockApi
, and replacing the current Bedrock ChatModelApi
with the newΒ BedrockConverseApi
.
from spring-ai.
@Hakenadu Great idea, I'll add this feature to PR's todo list. Also if you have any ideas plz let me know.
from spring-ai.
Hi @Hakenadu could you show some scenarios to explain why you would use the Amazon Bedrock Converse API
directly?
If you just use the converse you still need to pay attention to managing the parameters with different models, I'm thinking it might be best to keep current Spring-AI ways to call different models and change the native AWS BedrockRuntime SDK
to use Bedrock Converse API
from spring-ai.
Hi @maxjiang153, we are currently implementing an API that allows access to different models from various providers (of which AWS Bedrock is just one), abstracting away the model choice for an end user.
By implementing the Converse API, I hope to have a more manageable set of beans (especially since we want to customise models backwards).
On the other hand, I can well imagine exposing the Converse API format, which is designed for abstraction, to the outside world. Specifically, I would try to reuse the model classes in our own API. In fact, I am currently doing this with the (Spring AI) OpenAI request models.
Of course, it may be that these are very specific framework conditions. Nevertheless, I believe that the Converse API can be a good addition to the Bedrock implementations in Spring AI.
from spring-ai.
@maxjiang153 I already took a look into you changes, thank you for your work in this matter!!
I noticed that (also in the current implementation) it isn't possible to pass a custom BedrockRuntimeClient to the AbstractBedrockApi. Could you perhaps also directly include a corresponding constructor for this?
that would be great!
from spring-ai.
Related Issues (20)
- Create an AbstractChatModel class that adds logging of request and response. HOT 1
- Duplicate index handling when using Spring AI and MongoDB
- Add ChatClient Logging Advisor
- Add support for Amazon Bedrock Claude 3.5 Sonnet
- Hope to increase custom host function HOT 3
- The company has deployed a large open source model and provided an http interface. Now how can I use Spring AI to access the http interface of the private model? HOT 1
- Function Calling with Kotlin Functions HOT 1
- Unstable Testcontainers tests
- Remove PaLM Model support as it is deprecated and will be decomissioned
- Provide Anthropic Claude streaming Function Calling support HOT 2
- Add comment in OpenAiAudioSpeechModel and remove duplicated comment in OpenAiAudioSpeechOptions
- temperature and topP for Options should be Double, not Float
- Include "id" and "model" in ChatResponseMetadata for better traceability
- Add Couchbase Vector Store integration
- PgVector column cannot have more than 2000 dimensions for index HOT 2
- Concurrent Function/Tool Invocation
- VertexAiGeminiChatModel: Make safetySettings configurable
- Prompt template validation failed due to the variable name contains hyphen(-) symbol
- Suggestion: StructuredOutputConverter implementations could have configured amount of retries when convert() is failing to parse response HOT 1
- [Observability] Initial Observability for Chat Models
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. πππ
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from spring-ai.