diff --git a/README.md b/README.md index 11475da8f..80f1d437a 100644 --- a/README.md +++ b/README.md @@ -16,18 +16,17 @@ ## What is Arcadia? -**Arcadia** comes from [Greek mythology](https://www.greekmythology.com/Myths/Places/Arcadia/arcadia.html)(a tranquil and idyllic region, representing harmony, serenity, and natural beauty). We aim to help everyone find a more perfect integration between humans and AI. - -To achieve this goal, we provide this one-stop LLMOps solution. Furthermore, we can easily host **Arcadia** at any Kubernetes cluster as production ready by integrating [kubebb](https://github.com/kubebb)(Kubernetes building blocks). +**Arcadia** is a one-stop enterprise-grade LLMOps platform that provides a unified interface for developers and operators to build, debug,deploy and manage AI agents with a orchestration engine(**RAG(Retrieval Augmented Generation)** and **LLM finetuning**). ## Features -* Multi-tenant isolation (data, model services), built-in OIDC, RBAC, and auditing, supporting different companies and departments to develop through a unified platform -* Kubernetes native AGI agent orchestration + +* Build,debug,deploy AI agents on ops-console(GUI for LLMOps) +* Chat with AGI agent on agent-portal(GUI for gpt chat) +* Enterprise-grade infratructure with [KubeBB](https://github.com/kubebb): Multi-tenant isolation (data, model services), built-in OIDC, RBAC, and auditing, supporting different companies and departments to develop through a unified platform +* Support most of the popular LLMs(large language models),embedding models,reranking models,etc.. +* Inference acceleration with [vllm](https://github.com/vllm-project/vllm),distributed inference with [ray](https://github.com/ray-project/ray),quantization, and more +* Support fine-tuining with [llama-factory](https://github.com/hiyouga/LLaMA-Factory) * Built on langchaingo(golang), has better performance and maintainability -* Support distributed inference using Ray -* Support quality and performance evaluation of AGI agent under different configurations -* A development and operational platform for AI agents, along with an AI agent portal for end-users -* Developed based on micro frontends and low-code approach, allowing for quick scalability and integration ## Architecture @@ -45,52 +44,59 @@ Visit our [online documents](http://kubeagi.k8s.com.cn/docs/intro) Read [user guide](http://kubeagi.k8s.com.cn/docs/UserGuide/intro) +## Supported Models + +### List of Models can be deployed by kubeagi + ### LLMs -List of supported(tested) LLMs -* baichuan2-7b -* chatglm2-6b -* qwen-7b-chat / qwen-14b-chat / qwen-72b-chat -* llama2-7b -* Mistral-7B-Instruct-v0.1 -* bge-large-zh ***embedding*** -* m3e ***embedding*** -* [ZhiPuAI(智谱 AI)](https://github.com/kubeagi/arcadia/tree/main/pkg/llms/zhipuai) - - [example](https://github.com/kubeagi/arcadia/blob/main/examples/zhipuai/main.go) - - [embedding](https://github.com/kubeagi/arcadia/tree/main/pkg/embeddings/zhipuai) -* [DashScope(灵积模型服务)](https://github.com/kubeagi/arcadia/tree/main/pkg/llms/dashscope) - - [example](https://github.com/kubeagi/arcadia/blob/main/examples/dashscope/main.go) - - [text-embedding-v1(通用文本向量 同步接口)](https://help.aliyun.com/zh/dashscope/developer-reference/text-embedding-api-details) +* [chatglm2-6b](https://huggingface.co/THUDM/chatglm2-6b) +* [chatglm3-6b](https://huggingface.co/THUDM/chatglm3-6b>) +* [qwen(7B,14B,72B)](https://huggingface.co/Qwen) +* [qwen-1.5(0.5B,1.8B,4B,14B,32B](https://huggingface.co/collections/Qwen/qwen15-65c0a2f577b1ecb76d786524) +* [baichuan2](https://huggingface.co/baichuan-inc) +* [llama2](https://huggingface.co/meta-llama) +* [mistral](https://huggingface.co/mistralai) ### Embeddings -> Fully compatible with [langchain embeddings](https://github.com/tmc/langchaingo/tree/main/embeddings) +* [bge-large-zh](https://huggingface.co/BAAI/bge-large-zh-v1.5) +* [m3e](https://huggingface.co/moka-ai/m3e-base) + +### Reranking + +* [bge-reranker-large](https://huggingface.co/BAAI/bge-reranker-large) ***reranking*** +* [bce-reranking]() ***reranking*** -### VectorStores +### List of Online(third party) LLM Services can be integrated by kubeagi + +* [OpenAI](https://openai.com/) +* [Google Gemini](https://gemini.google.com/) +* [智谱AI](https://github.com/kubeagi/arcadia/tree/main/pkg/llms/zhipuai) + * [example](https://github.com/kubeagi/arcadia/blob/main/examples/zhipuai/main.go) + * [embedding](https://github.com/kubeagi/arcadia/tree/main/pkg/embeddings/zhipuai) +* [DashScope(灵积模型服务)](https://github.com/kubeagi/arcadia/tree/main/pkg/llms/dashscope) + * [example](https://github.com/kubeagi/arcadia/blob/main/examples/dashscope/main.go) + * [text-embedding-v1(通用文本向量 同步接口)](https://help.aliyun.com/zh/dashscope/developer-reference/text-embedding-api-details) + +## Supported VectorStores > Fully compatible with [langchain vectorstores](https://github.com/tmc/langchaingo/tree/main/vectorstores) -- ✅ [PG Vector](https://github.com/tmc/langchaingo/tree/main/vectorstores/pgvector), KubeAGI adds the PG vector support to [langchaingo](https://github.com/tmc/langchaingo) project. -- ✅ [ChromaDB](https://docs.trychroma.com/) +* ✅ [PG Vector](https://github.com/tmc/langchaingo/tree/main/vectorstores/pgvector), KubeAGI adds the PG vector support to [langchaingo](https://github.com/tmc/langchaingo) project. +* ✅ [ChromaDB](https://docs.trychroma.com/) ## Pure Go Toolchains Thanks to [langchaingo](https://github.com/tmc/langchaingo),we can have comprehensive AI capability in Golang!But in order to meet our own unique needs, we have further developed a number of other toolchains: -- [Optimized DocumentLoaders](https://github.com/kubeagi/arcadia/tree/main/pkg/documentloaders): optimized csv,etc... -- [Extended LLMs](https://github.com/kubeagi/arcadia/tree/main/pkg/llms): zhipuai,dashscope,etc... -- [Tools](https://github.com/kubeagi/arcadia/tree/main/pkg/tools): bingsearch,weather,etc... -- [AppRuntime](https://github.com/kubeagi/arcadia/tree/main/pkg/appruntime): powerful node(LLM,Chain,KonwledgeBase,vectorstore,Agent,etc...) orchestration runtime for arcadia +* [Optimized DocumentLoaders](https://github.com/kubeagi/arcadia/tree/main/pkg/documentloaders): optimized csv,etc... +* [Extended LLMs](https://github.com/kubeagi/arcadia/tree/main/pkg/llms): zhipuai,dashscope,etc... +* [Tools](https://github.com/kubeagi/arcadia/tree/main/pkg/tools): bingsearch,weather,etc... +* [AppRuntime](https://github.com/kubeagi/arcadia/tree/main/pkg/appruntime): powerful node(LLM,Chain,KonwledgeBase,vectorstore,Agent,etc...) orchestration runtime for arcadia We have provided some examples on how to use them. See more details at [here](https://github.com/kubeagi/arcadia/tree/main/examples) -## CLI - -We provide a Command Line Tool `arctl` to interact with `arcadia`. See [here](http://kubeagi.k8s.com.cn/docs/Tools/arctl-tool) for more details. - -- ✅ datasource management -- ✅ RAG evaluation - ## Contribute to Arcadia If you want to contribute to Arcadia, refer to [contribute guide](http://kubeagi.k8s.com.cn/docs/Contribute/prepare-and-start). diff --git a/deploy/charts/arcadia/Chart.yaml b/deploy/charts/arcadia/Chart.yaml index 4f25507cf..3c17857a9 100644 --- a/deploy/charts/arcadia/Chart.yaml +++ b/deploy/charts/arcadia/Chart.yaml @@ -3,7 +3,7 @@ name: arcadia description: A Helm chart(Also a KubeBB Component) for KubeAGI Arcadia type: application version: 0.3.30 -appVersion: "0.2.1" +appVersion: "0.2.2" keywords: - LLMOps diff --git a/deploy/charts/arcadia/values.yaml b/deploy/charts/arcadia/values.yaml index 66c7b8f04..cfb5b4613 100644 --- a/deploy/charts/arcadia/values.yaml +++ b/deploy/charts/arcadia/values.yaml @@ -37,7 +37,7 @@ config: controller: # 1: error 3:info 5:debug loglevel: 3 - image: kubeagi/arcadia:v0.2.1-20240401-b80e4e4 + image: kubeagi/arcadia:v0.2.2 imagePullPolicy: IfNotPresent resources: limits: @@ -51,7 +51,7 @@ controller: # related project: https://github.com/kubeagi/arcadia/tree/main/apiserver apiserver: loglevel: 3 - image: kubeagi/arcadia:v0.2.1-20240401-b80e4e4 + image: kubeagi/arcadia:v0.2.2 enableplayground: false port: 8081 ingress: @@ -70,7 +70,7 @@ apiserver: opsconsole: enabled: true kubebbEnabled: true - image: kubeagi/ops-console:v0.2.1-20240401-2e63d80 + image: kubeagi/ops-console:v0.2.2 ingress: path: kubeagi-portal-public host: portal..nip.io @@ -81,7 +81,7 @@ gpts: # all gpt resources are public in this namespace public_namespace: gpts agentportal: - image: kubeagi/agent-portal:v0.1.0-20240401-bc9e42d + image: kubeagi/agent-portal:v0.1.0-20240411-e26a310 ingress: path: "" host: gpts..nip.io @@ -91,7 +91,7 @@ fastchat: enabled: true image: repository: kubeagi/arcadia-fastchat - tag: v0.2.36 + tag: v0.2.36-patch ingress: enabled: true host: fastchat-api..nip.io @@ -131,7 +131,7 @@ minio: # Related project: https://github.com/kubeagi/arcadia/tree/main/data-processing dataprocess: enabled: true - image: kubeagi/data-processing:v0.2.1 + image: kubeagi/data-processing:v0.2.2 port: 28888 config: llm: