From 6f7625fa47225b41b515137a6f6a2caea959a705 Mon Sep 17 00:00:00 2001
From: Aaron Ji <127167174+DresAaron@users.noreply.github.com>
Date: Sat, 14 Sep 2024 16:21:17 +0800
Subject: [PATCH] chore: update Jina embedding model (#8376)

---
 api/core/model_runtime/model_providers/jina/jina.py      | 4 ++--
 .../jina/text_embedding/jina-embeddings-v3.yaml          | 9 +++++++++
 .../jina/text_embedding/text_embedding.py                | 3 +++
 3 files changed, 14 insertions(+), 2 deletions(-)
 create mode 100644 api/core/model_runtime/model_providers/jina/text_embedding/jina-embeddings-v3.yaml

diff --git a/api/core/model_runtime/model_providers/jina/jina.py b/api/core/model_runtime/model_providers/jina/jina.py
index 33977b6a33..186a0a0fa7 100644
--- a/api/core/model_runtime/model_providers/jina/jina.py
+++ b/api/core/model_runtime/model_providers/jina/jina.py
@@ -18,9 +18,9 @@ class JinaProvider(ModelProvider):
         try:
             model_instance = self.get_model_instance(ModelType.TEXT_EMBEDDING)
 
-            # Use `jina-embeddings-v2-base-en` model for validate,
+            # Use `jina-embeddings-v3` model for validate,
             # no matter what model you pass in, text completion model or chat model
-            model_instance.validate_credentials(model="jina-embeddings-v2-base-en", credentials=credentials)
+            model_instance.validate_credentials(model="jina-embeddings-v3", credentials=credentials)
         except CredentialsValidateFailedError as ex:
             raise ex
         except Exception as ex:
diff --git a/api/core/model_runtime/model_providers/jina/text_embedding/jina-embeddings-v3.yaml b/api/core/model_runtime/model_providers/jina/text_embedding/jina-embeddings-v3.yaml
new file mode 100644
index 0000000000..4e5374dc9d
--- /dev/null
+++ b/api/core/model_runtime/model_providers/jina/text_embedding/jina-embeddings-v3.yaml
@@ -0,0 +1,9 @@
+model: jina-embeddings-v3
+model_type: text-embedding
+model_properties:
+  context_size: 8192
+  max_chunks: 2048
+pricing:
+  input: '0.001'
+  unit: '0.001'
+  currency: USD
diff --git a/api/core/model_runtime/model_providers/jina/text_embedding/text_embedding.py b/api/core/model_runtime/model_providers/jina/text_embedding/text_embedding.py
index ef12e534db..5033f0f748 100644
--- a/api/core/model_runtime/model_providers/jina/text_embedding/text_embedding.py
+++ b/api/core/model_runtime/model_providers/jina/text_embedding/text_embedding.py
@@ -56,6 +56,9 @@ class JinaTextEmbeddingModel(TextEmbeddingModel):
 
         data = {"model": model, "input": [transform_jina_input_text(model, text) for text in texts]}
 
+        if model == "jina-embeddings-v3":
+            data["task_type"] = "retrieval.passage"
+
         try:
             response = post(url, headers=headers, data=dumps(data))
         except Exception as e: