docs: update the LLM / Embedding models for Gitee AI (#551)

close #549 --------- Co-authored-by: Mini256 <minianter@foxmail.com>
pingcap · Dec 29, 2024 · 9b05f10 · 9b05f10
1 parent 12aa507
commit 9b05f10
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 1 deletion.
diff --git a/backend/app/rag/embed_model_option.py b/backend/app/rag/embed_model_option.py
@@ -91,7 +91,7 @@ class EmbeddingModelOption(BaseModel):
         provider_description="Gitee AI is a third-party model provider that offers ready-to-use cutting-edge model APIs for AI developers.",
         provider_url="https://ai.gitee.com",
         default_embedding_model="bge-large-zh-v1.5",
-        embedding_model_description=f"Find more information about Gitee AI Embeddings at https://ai.gitee.com/docs/openapi/v1#tag/serverless/POST/embeddings",
+        embedding_model_description=f"Find more information about Gitee AI Embeddings at https://ai.gitee.com/docs/openapi/v1#tag/%E7%89%B9%E5%BE%81%E6%8A%BD%E5%8F%96/POST/embeddings",
         credentials_display_name="Gitee AI API Key",
         credentials_description="The API key of Gitee AI, you can find it in https://ai.gitee.com/dashboard/settings/tokens",
         credentials_type="str",

diff --git a/frontend/app/src/pages/docs/embedding-model.mdx b/frontend/app/src/pages/docs/embedding-model.mdx
@@ -130,6 +130,20 @@ To use Ollama, you'll need to configure the API base URL in the **Advanced Setti
 
 For more information, see the [Ollama embedding models documentation](https://ollama.com/blog/embedding-models).
 
+### Gitee AI
+
+Gitee AI is a third-party model provider that offers ready-to-use cutting-edge model APIs for AI developers.
+
+**Supported Models**:
+
+| Embedding Model     | Vector Dimensions | Max Tokens |
+| ------------------- | ----------------- | ---------- |
+| `bge-m3`            | 1024              | 8192       |
+| `bge-large-zh-v1.5` | 1024              | 512        |
+| `bge-small-zh-v1.5` | 512               | 512        |
+
+For more information, see the [Gitee AI embedding models documentation](https://ai.gitee.com/docs/openapi/v1#tag/%E7%89%B9%E5%BE%81%E6%8A%BD%E5%8F%96/POST/embeddings).
+
 ### Local Embedding Server
 
 Autoflow's local embedding server is a self-hosted embedding service built upon [sentence-transformers](https://www.sentence-transformers.org/) and deployed on your own infrastructure.

diff --git a/frontend/app/src/pages/docs/llm.mdx b/frontend/app/src/pages/docs/llm.mdx
@@ -30,6 +30,7 @@ Currently Autoflow supports the following LLM providers:
 - [Google Gemini](https://gemini.google.com/)
 - [Anthropic Vertex AI](https://cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude)
 - [Amazon Bedrock](https://aws.amazon.com/bedrock/)
+- [Gitee AI](https://ai.gitee.com/serverless-api)
 - And all OpenAI-Like models:
     - [OpenRouter](https://openrouter.ai/)
         - Default config: