fix: jina embedding error (#5839)

* fix: jina embedding error

* fix: handle jina model encoding format in Embeddings class

.yarn/patches/openai-npm-4.96.0-0665b05cb9.patch

@@ -37,3 +37,49 @@ index 7d19f5578040afa004bc887aab1725e8703d2bac..59ec725b6142299a62798ac4bdedb63b
          if (status === 400) {
              return new BadRequestError(status, error, message, headers);
          }
+diff --git a/resources/embeddings.js b/resources/embeddings.js
+index aae578404cb2d09a39ac33fc416f1c215c45eecd..25c54b05bdae64d5c3b36fbb30dc7c8221b14034 100644
+--- a/resources/embeddings.js
++++ b/resources/embeddings.js
+@@ -36,6 +36,9 @@ class Embeddings extends resource_1.APIResource {
+         // No encoding_format specified, defaulting to base64 for performance reasons
+         // See https://github.com/openai/openai-node/pull/1312
+         let encoding_format = hasUserProvidedEncodingFormat ? body.encoding_format : 'base64';
++        if (body.model.includes('jina')) {
++            encoding_format = undefined;
++        }
+         if (hasUserProvidedEncodingFormat) {
+             Core.debug('Request', 'User defined encoding_format:', body.encoding_format);
+         }
+@@ -47,7 +50,7 @@ class Embeddings extends resource_1.APIResource {
+             ...options,
+         });
+         // if the user specified an encoding_format, return the response as-is
+-        if (hasUserProvidedEncodingFormat) {
++        if (hasUserProvidedEncodingFormat || body.model.includes('jina')) {
+             return response;
+         }
+         // in this stage, we are sure the user did not specify an encoding_format
+diff --git a/resources/embeddings.mjs b/resources/embeddings.mjs
+index 0df3c6cc79a520e54acb4c2b5f77c43b774035ff..aa488b8a11b2c413c0a663d9a6059d286d7b5faf 100644
+--- a/resources/embeddings.mjs
++++ b/resources/embeddings.mjs
+@@ -10,6 +10,9 @@ export class Embeddings extends APIResource {
+         // No encoding_format specified, defaulting to base64 for performance reasons
+         // See https://github.com/openai/openai-node/pull/1312
+         let encoding_format = hasUserProvidedEncodingFormat ? body.encoding_format : 'base64';
++        if (body.model.includes('jina')) {
++            encoding_format = undefined;
++        }
+         if (hasUserProvidedEncodingFormat) {
+             Core.debug('Request', 'User defined encoding_format:', body.encoding_format);
+         }
+@@ -21,7 +24,7 @@ export class Embeddings extends APIResource {
+             ...options,
+         });
+         // if the user specified an encoding_format, return the response as-is
+-        if (hasUserProvidedEncodingFormat) {
++        if (hasUserProvidedEncodingFormat || body.model.includes('jina')) {
+             return response;
+         }
+         // in this stage, we are sure the user did not specify an encoding_format

src/renderer/src/pages/knowledge/components/AddKnowledgePopup.tsx

@@ -14,7 +14,7 @@ import { getErrorMessage } from '@renderer/utils/error'
 import { Form, Input, Modal, Select, Slider } from 'antd'
 import { find, sortBy } from 'lodash'
 import { nanoid } from 'nanoid'
-import { useRef, useState } from 'react'
+import { useMemo, useRef, useState } from 'react'
 import { useTranslation } from 'react-i18next'
 
 interface ShowParams {
@@ -40,59 +40,69 @@ const PopupContainer: React.FC<Props> = ({ title, resolve }) => {
   const { providers } = useProviders()
   const { addKnowledgeBase } = useKnowledgeBases()
 
-  const allModels = providers
+  const embeddingModels = useMemo(() => {
-    .map((p) => p.models)
+    return providers
-    .flat()
+      .map((p) => p.models)
-    .filter((model) => isEmbeddingModel(model) && !isRerankModel(model))
+      .flat()
+      .filter((model) => isEmbeddingModel(model))
+  }, [providers])
 
-  const rerankModels = providers
+  const rerankModels = useMemo(() => {
-    .map((p) => p.models)
+    return providers
-    .flat()
+      .map((p) => p.models)
-    .filter((model) => isRerankModel(model))
+      .flat()
+      .filter((model) => isRerankModel(model))
+  }, [providers])
 
   const nameInputRef = useRef<any>(null)
 
-  const selectOptions = providers
+  const embeddingSelectOptions = useMemo(() => {
-    .filter((p) => p.models.length > 0)
+    console.log(providers)
-    .map((p) => ({
+    return providers
-      label: p.isSystem ? t(`provider.${p.id}`) : p.name,
+      .filter((p) => p.models.length > 0)
-      title: p.name,
+      .map((p) => ({
-      options: sortBy(p.models, 'name')
+        label: p.isSystem ? t(`provider.${p.id}`) : p.name,
-        .filter((model) => isEmbeddingModel(model) && !isRerankModel(model))
+        title: p.name,
-        .map((m) => ({
+        options: sortBy(p.models, 'name')
-          label: m.name,
+          .filter((model) => isEmbeddingModel(model))
-          value: getModelUniqId(m)
+          .map((m) => ({
-        }))
+            label: m.name,
-    }))
+            value: getModelUniqId(m),
-    .filter((group) => group.options.length > 0)
+            key: `${p.id}-${m.id}`
+          }))
+      }))
+      .filter((group) => group.options.length > 0)
+  }, [providers, t])
 
-  const rerankSelectOptions = providers
+  const rerankSelectOptions = useMemo(() => {
-    .filter((p) => p.models.length > 0)
+    return providers
-    .filter((p) => !NOT_SUPPORTED_REANK_PROVIDERS.includes(p.id))
+      .filter((p) => p.models.length > 0)
-    .map((p) => ({
+      .filter((p) => !NOT_SUPPORTED_REANK_PROVIDERS.includes(p.id))
-      label: p.isSystem ? t(`provider.${p.id}`) : p.name,
+      .map((p) => ({
-      title: p.name,
+        label: p.isSystem ? t(`provider.${p.id}`) : p.name,
-      options: sortBy(p.models, 'name')
+        title: p.name,
-        .filter((model) => isRerankModel(model))
+        options: sortBy(p.models, 'name')
-        .map((m) => ({
+          .filter((model) => isRerankModel(model))
-          label: m.name,
+          .map((m) => ({
-          value: getModelUniqId(m)
+            label: m.name,
-        }))
+            value: getModelUniqId(m)
-    }))
+          }))
-    .filter((group) => group.options.length > 0)
+      }))
+      .filter((group) => group.options.length > 0)
+  }, [providers, t])
 
   const onOk = async () => {
     try {
       const values = await form.validateFields()
-      const selectedModel = find(allModels, JSON.parse(values.model)) as Model
+      const selectedEmbeddingModel = find(embeddingModels, JSON.parse(values.model)) as Model
 
       const selectedRerankModel = values.rerankModel
         ? (find(rerankModels, JSON.parse(values.rerankModel)) as Model)
         : undefined
 
-      if (selectedModel) {
+      if (selectedEmbeddingModel) {
         setLoading(true)
-        const provider = providers.find((p) => p.id === selectedModel.provider)
+        const provider = providers.find((p) => p.id === selectedEmbeddingModel.provider)
 
         if (!provider) {
           return
@@ -102,7 +112,7 @@ const PopupContainer: React.FC<Props> = ({ title, resolve }) => {
         let dimensions = 0
 
         try {
-          dimensions = await aiProvider.getEmbeddingDimensions(selectedModel)
+          dimensions = await aiProvider.getEmbeddingDimensions(selectedEmbeddingModel)
         } catch (error) {
           console.error('Error getting embedding dimensions:', error)
           window.message.error(t('message.error.get_embedding_dimensions') + '\n' + getErrorMessage(error))
@@ -113,7 +123,7 @@ const PopupContainer: React.FC<Props> = ({ title, resolve }) => {
         const newBase = {
           id: nanoid(),
           name: values.name,
-          model: selectedModel,
+          model: selectedEmbeddingModel,
           rerankModel: selectedRerankModel,
           dimensions,
           documentCount: values.documentCount || DEFAULT_KNOWLEDGE_DOCUMENT_COUNT,
@@ -166,7 +176,7 @@ const PopupContainer: React.FC<Props> = ({ title, resolve }) => {
           label={t('models.embedding_model')}
           tooltip={{ title: t('models.embedding_model_tooltip'), placement: 'right' }}
           rules={[{ required: true, message: t('message.error.enter.model') }]}>
-          <Select style={{ width: '100%' }} options={selectOptions} placeholder={t('settings.models.empty')} />
+          <Select style={{ width: '100%' }} options={embeddingSelectOptions} placeholder={t('settings.models.empty')} />
         </Form.Item>
 
         <Form.Item

yarn.lock

@@ -12940,7 +12940,7 @@ __metadata:
 
 "openai@patch:openai@npm%3A4.96.0#~/.yarn/patches/openai-npm-4.96.0-0665b05cb9.patch":
   version: 4.96.0
-  resolution: "openai@patch:openai@npm%3A4.96.0#~/.yarn/patches/openai-npm-4.96.0-0665b05cb9.patch::version=4.96.0&hash=17e659"
+  resolution: "openai@patch:openai@npm%3A4.96.0#~/.yarn/patches/openai-npm-4.96.0-0665b05cb9.patch::version=4.96.0&hash=645779"
   dependencies:
     "@types/node": "npm:^18.11.18"
     "@types/node-fetch": "npm:^2.6.4"
@@ -12959,7 +12959,7 @@ __metadata:
       optional: true
   bin:
     openai: bin/cli
-  checksum: 10c0/b1f6162017ede2e0c3338ca94ea0e0c6ababc39ef8abea9e1a04d747725cf6ca3fbd0e4682c231af03a6473228b25a16ce52aac03c3cc4feb302d03b9603e06b
+  checksum: 10c0/8c16fcf1812294220eddd4616e298c2af87398acb479287b7565548c8c1979c6d5c487fb7a9c25b0ac59f778de74c23d94ce1a34362c49260ae7a14acf22abc2
   languageName: node
   linkType: hard