Merge pull request #171 from n4ze3m/next

v1.2.1

package.json

@@ -77,7 +77,7 @@
     "tailwindcss": "^3.4.1",
     "typescript": "5.3.3",
     "vite-plugin-top-level-await": "^1.4.1",
-    "wxt": "^0.17.7"
+    "wxt": "^0.19.6"
   },
   "resolutions": {
     "@langchain/core": "0.1.45"

src/assets/locale/en/common.json

@@ -81,6 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "Enter Top P value (e.g. 0.9, 0.95)"
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "Enter number of layers to send to GPU(s)"
             }
         },
         "advanced": "More Model Settings"

src/assets/locale/en/settings.json

@@ -308,9 +308,14 @@
         "required": "Please enter a chunk overlap"
       },
       "totalFilePerKB": {
-        "label": "Knowledge Base Default File Limit",
+        "label": "Knowledge Base Default File Upload Limit",
-        "placeholder": "Enter default file limit (e.g., 10)",
+        "placeholder": "Enter default file upload limit (e.g., 10)",
-        "required": "Please enter the default file limit"
+        "required": "Please enter the default file upload limit"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Number of Retrieved Documents",
+        "placeholder": "Enter Number of Retrieved Documents",
+        "required": "Please enter the number of retrieved documents"
       }
     },
     "prompt": {

src/assets/locale/es/common.json

@@ -81,6 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "Ingresar el valor de Top P (ej: 0.9, 0.95)"
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "Ingrese el número de capas para enviar a la(s) GPU(s)"
             }
         },
         "advanced": "Más Configuraciones del Modelo"

src/assets/locale/es/settings.json

@@ -30,7 +30,7 @@
       "sendNotificationAfterIndexing": {
         "label": "Enviar notificación después de terminar el procesamiento de la base de conocimientos"
       },
-      "generateTitle" :{
+      "generateTitle": {
         "label": "Generar título usando IA"
       }
     },
@@ -308,9 +308,14 @@
         "required": "Por favor, ingresar el solapamiento del chunk"
       },
       "totalFilePerKB": {
-        "label": "Límite predeterminado de archivos de la base de conocimientos",
+        "label": "Límite predeterminado de carga de archivos para la Base de Conocimientos",
-        "placeholder": "Ingrese el límite predeterminado de archivos (ej. 10)",
+        "placeholder": "Ingrese el límite predeterminado de carga de archivos (ej., 10)",
-        "required": "Por favor, ingrese el límite predeterminado de archivos"
+        "required": "Por favor, ingrese el límite predeterminado de carga de archivos"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Número de Documentos Recuperados",
+        "placeholder": "Ingrese el Número de Documentos Recuperados",
+        "required": "Por favor, ingrese el número de documentos recuperados"
       }
     },
     "prompt": {

src/assets/locale/fa/common.json

@@ -81,8 +81,12 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "مقدار Top P را وارد کنید (مثلا 0.9, 0.95)"
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "تعداد لایه‌هایی که به GPU(ها) ارسال می‌شود را وارد کنید"
             }
         },
         "advanced": "تنظیمات بیشتر مدل"
     }
-}
+}

src/assets/locale/fa/settings.json

@@ -30,7 +30,7 @@
       "sendNotificationAfterIndexing": {
         "label": "ارسال نوتیفیکیشن پس از اتمام پردازش پایگاه دانش"
       },
-      "generateTitle" :{
+      "generateTitle": {
         "label": "تولید عنوان با استفاده از هوش مصنوعی"
       }
     },
@@ -307,9 +307,14 @@
         "required": "لطفا یک همپوشانی تکه ای وارد کنید"
       },
       "totalFilePerKB": {
-        "label": "محدودیت فایل پیش فرض پایگاه دانش",
+        "label": "محدودیت پیش‌فرض آپلود فایل پایگاه دانش",
-        "placeholder": "محدودیت فایل پیش فرض را وارد کنید (به عنوان مثال، 10)",
+        "placeholder": "محدودیت پیش‌فرض آپلود فایل را وارد کنید (مثلاً 10)",
-        "required": "لطفا محدودیت پیش فرض فایل را وارد کنید"
+        "required": "لطفاً محدودیت پیش‌فرض آپلود فایل را وارد کنید"
+      },
+      "noOfRetrievedDocs": {
+        "label": "تعداد اسناد بازیابی شده",
+        "placeholder": "تعداد اسناد بازیابی شده را وارد کنید",
+        "required": "لطفاً تعداد اسناد بازیابی شده را وارد کنید"
       }
     },
     "prompt": {

src/assets/locale/fr/common.json

@@ -81,6 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "Entrez la valeur Top P (par exemple 0,9, 0,95)"
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "Entrez le nombre de couches à envoyer au(x) GPU(s)"
             }
         },
         "advanced": "Plus de paramètres du modèle"

src/assets/locale/fr/settings.json

@@ -29,8 +29,8 @@
       },
       "sendNotificationAfterIndexing": {
         "label": "Envoyer une notification après avoir terminé le traitement de la base de connaissances"
-      }, 
+      },
-      "generateTitle" :{
+      "generateTitle": {
         "label": "Générer le titre en utilisant l'IA"
       }
     },
@@ -308,9 +308,14 @@
         "required": "Veuillez saisir un chevauchement"
       },
       "totalFilePerKB": {
-        "label": "Limite par défaut de fichiers de la base de connaissances",
+        "label": "Limite par défaut de téléchargement de fichiers pour la base de connaissances",
-        "placeholder": "Entrez la limite par défaut de fichiers (ex. 10)",
+        "placeholder": "Entrez la limite par défaut de téléchargement de fichiers (par exemple, 10)",
-        "required": "Veuillez entrer la limite par défaut de fichiers"
+        "required": "Veuillez saisir la limite par défaut de téléchargement de fichiers"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Nombre de documents récupérés",
+        "placeholder": "Entrez le nombre de documents récupérés",
+        "required": "Veuillez saisir le nombre de documents récupérés"
       }
     },
     "prompt": {

src/assets/locale/it/common.json

@@ -81,7 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "Inserisci il Valore Top P (e.g. 0.9, 0.95)"
-            }
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "Inserisci il numero di layer da inviare alla/e GPU"            }
         },
         "advanced": "Altre Impostazioni del Modello"
     },

src/assets/locale/it/settings.json

@@ -29,8 +29,8 @@
       },
       "sendNotificationAfterIndexing": {
         "label": "Inviare notifica dopo aver terminato l'elaborazione della base di conoscenza"
-      }, 
+      },
-      "generateTitle" :{
+      "generateTitle": {
         "label": "Genera titolo utilizzando l'IA"
       }
     },
@@ -308,9 +308,14 @@
         "required": "Inserisci la Sovrapposizione del Blocco"
       },
       "totalFilePerKB": {
-        "label": "Limite predefinito di file della base di conoscenza",
+        "label": "Limite Predefinito di Caricamento File per la Base di Conoscenza",
-        "placeholder": "Inserisci il limite predefinito di file (es. 10)",
+        "placeholder": "Inserisci il limite predefinito di caricamento file (es. 10)",
-        "required": "Inserisci il limite predefinito di file"
+        "required": "Inserisci il limite predefinito di caricamento file"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Numero di Documenti Recuperati",
+        "placeholder": "Inserisci il Numero di Documenti Recuperati",
+        "required": "Inserisci il numero di documenti recuperati"
       }
     },
     "prompt": {

src/assets/locale/ja-JP/common.json

@@ -81,6 +81,10 @@
       "topP": {
         "label": "Top P",
         "placeholder": "Top P値を入力してください（例：0.9、0.95）"
+      },
+      "numGpu": {
+        "label": "Num GPU",
+        "placeholder": "GPU(s)に送信するレイヤー数を入力してください"
       }
     },
     "advanced": "その他のモデル設定"

src/assets/locale/ja-JP/settings.json

@@ -311,9 +311,14 @@
         "required": "チャンクオーバーラップを入力してください"
       },
       "totalFilePerKB": {
-        "label": "ナレッジベースのデフォルトファイル制限",
+        "label": "知識ベースのデフォルトファイルアップロード制限",
-        "placeholder": "デフォルトのファイル制限を入力してください（例：10）",
+        "placeholder": "デフォルトのファイルアップロード制限を入力してください（例：10）",
-        "required": "デフォルトのファイル制限を入力してください"
+        "required": "デフォルトのファイルアップロード制限を入力してください"
+      },
+      "noOfRetrievedDocs": {
+        "label": "取得ドキュメント数",
+        "placeholder": "取得ドキュメント数を入力",
+        "required": "取得ドキュメント数を入力してください"
       }
     },
     "prompt": {

src/assets/locale/ml/common.json

@@ -80,6 +80,10 @@
             "topP": {
                 "label": "ടോപ് P",
                 "placeholder": "ടോപ് P മൂല്യം നൽകുക (ഉദാ: 0.9, 0.95)"
+            },
+            "numGpu": {
+                "label": "ജിപിയു എണ്ണം",
+                "placeholder": "ജിപിയു(കൾ)ക്ക് അയക്കേണ്ട ലേയറുകളുടെ എണ്ണം നൽകുക"
             }
         },
         "advanced": "കൂടുതൽ മോഡൽ ക്രമീകരണങ്ങൾ"

src/assets/locale/ml/settings.json

@@ -311,9 +311,14 @@
         "required": "ദയവായി ചങ്ക് ഓവര്‍ലാപ്പ് നല്കുക"
       },
       "totalFilePerKB": {
-        "label": "Limite padrão de arquivos da base de conhecimento",
+        "label": "വിജ്ഞാനാധാരത്തിന്റെ സ്ഥിര ഫയൽ അപ്‌ലോഡ് പരിധി",
-        "placeholder": "Digite o limite padrão de arquivos (ex. 10)",
+        "placeholder": "സ്ഥിര ഫയൽ അപ്‌ലോഡ് പരിധി നൽകുക (ഉദാ: 10)",
-        "required": "Por favor, digite o limite padrão de arquivos"
+        "required": "ദയവായി സ്ഥിര ഫയൽ അപ്‌ലോഡ് പരിധി നൽകുക"
+      },
+      "noOfRetrievedDocs": {
+        "label": "വീണ്ടെടുത്ത രേഖകളുടെ എണ്ണം",
+        "placeholder": "വീണ്ടെടുത്ത രേഖകളുടെ എണ്ണം നൽകുക",
+        "required": "ദയവായി വീണ്ടെടുത്ത രേഖകളുടെ എണ്ണം നൽകുക"
       }
     },
     "prompt": {

src/assets/locale/pt-BR/common.json

@@ -81,6 +81,10 @@
       "topP": {
         "label": "Top P",
         "placeholder": "Digite o valor do Top P (ex: 0.9, 0.95)"
+      },
+      "numGpu": {
+          "label": "Num GPUs",
+          "placeholder": "Digite o número de camadas para enviar para a(s) GPU(s)"
       }
     },
     "advanced": "Mais Configurações do Modelo"

src/assets/locale/pt-BR/settings.json

@@ -308,9 +308,14 @@
         "required": "Por favor, insira uma sobreposição de pedaço"
       },
       "totalFilePerKB": {
-        "label": "Limite padrão de arquivos da base de conhecimento",
+        "label": "Limite Padrão de Upload de Arquivos da Base de Conhecimento",
-        "placeholder": "Digite o limite padrão de arquivos (ex. 10)",
+        "placeholder": "Digite o limite padrão de upload de arquivos (ex: 10)",
-        "required": "Por favor, digite o limite padrão de arquivos"
+        "required": "Por favor, insira o limite padrão de upload de arquivos"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Número de Documentos Recuperados",
+        "placeholder": "Digite o Número de Documentos Recuperados",
+        "required": "Por favor, insira o número de documentos recuperados"
       }
     },
     "prompt": {

src/assets/locale/ru/common.json

@@ -81,7 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "Введите значение Top P (например, 0.9, 0.95)"
-            }
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "Введите количество слоев для отправки на GPU"            }
         },
         "advanced": "Больше настроек модели"
     },

src/assets/locale/ru/settings.json

@@ -310,9 +310,14 @@
         "required": "Пожалуйста, введите перекрытие фрагментов"
       },
       "totalFilePerKB": {
-        "label": "Стандартный лимит файлов базы знаний",
+        "label": "Лимит загрузки файлов по умолчанию для базы знаний",
-        "placeholder": "Введите стандартный лимит файлов (напр. 10)",
+        "placeholder": "Введите лимит загрузки файлов по умолчанию (например, 10)",
-        "required": "Пожалуйста, введите стандартный лимит файлов"
+        "required": "Пожалуйста, введите лимит загрузки файлов по умолчанию"
+      },
+      "noOfRetrievedDocs": {
+        "label": "Количество извлеченных документов",
+        "placeholder": "Введите количество извлеченных документов",
+        "required": "Пожалуйста, введите количество извлеченных документов"
       }
     },
     "prompt": {

src/assets/locale/zh/common.json

@@ -81,6 +81,10 @@
             "topP": {
                 "label": "Top P",
                 "placeholder": "输入Top P值（例如：0.9、0.95）"
+            },
+            "numGpu": {
+                "label": "Num GPU",
+                "placeholder": "输入要发送到 GPU 的层数"
             }
         },
         "advanced": "更多模型设置"

src/assets/locale/zh/settings.json

@@ -313,9 +313,14 @@
         "required": "请输入嵌入重叠"
       },
       "totalFilePerKB": {
-        "label": "知识库默认文件限制",
+        "label": "知识库默认文件上传限制",
-        "placeholder": "输入默认文件限制（例如：10）",
+        "placeholder": "输入默认文件上传限制（例如：10）",
-        "required": "请输入默认文件限制"
+        "required": "请输入默认文件上传限制"
+      },
+      "noOfRetrievedDocs": {
+        "label": "检索文档数量",
+        "placeholder": "输入检索文档数量",
+        "required": "请输入检索文档数量"
       }
     },
     "prompt": {

src/components/Common/Settings/CurrentChatModelSettings.tsx

@@ -24,7 +24,8 @@ export const CurrentChatModelSettings = ({ open, setOpen }: Props) => {
         topP: cUserSettings.topP ?? data.topP,
         keepAlive: cUserSettings.keepAlive ?? data.keepAlive,
         numCtx: cUserSettings.numCtx ?? data.numCtx,
-        seed: cUserSettings.seed
+        seed: cUserSettings.seed,
+        numGpu: cUserSettings.numGpu ?? data.numGpu,
       })
       return data
     },
@@ -118,6 +119,16 @@ export const CurrentChatModelSettings = ({ open, setOpen }: Props) => {
                         placeholder={t("modelSettings.form.topP.placeholder")}
                       />
                     </Form.Item>
+
+                    <Form.Item
+                      name="numGpu"
+                      label={t("modelSettings.form.numGpu.label")}>
+                      <InputNumber
+                        style={{ width: "100%" }}
+                        size="large"
+                        placeholder={t("modelSettings.form.numGpu.placeholder")}
+                      />
+                    </Form.Item>
                   </React.Fragment>
                 )
               }

src/components/Option/Settings/model-settings.tsx

@@ -6,7 +6,6 @@ import { Form, Skeleton, Input, InputNumber, Collapse } from "antd"
 import React from "react"
 import { useTranslation } from "react-i18next"
 
-
 export const ModelSettings = () => {
   const { t } = useTranslation("common")
   const [form] = Form.useForm()
@@ -24,10 +23,10 @@ export const ModelSettings = () => {
     <div>
       <div>
         <div className="inline-flex items-center gap-2">
-        <BetaTag />
+          <BetaTag />
-        <h2 className="text-base font-semibold leading-7 text-gray-900 dark:text-white">
+          <h2 className="text-base font-semibold leading-7 text-gray-900 dark:text-white">
-          {t("modelSettings.label")} 
+            {t("modelSettings.label")}
-        </h2>
+          </h2>
         </div>
         <p className="text-sm text-gray-700 dark:text-neutral-400 mt-1">
           {t("modelSettings.description")}
@@ -41,6 +40,7 @@ export const ModelSettings = () => {
             temperature: number
             topK: number
             topP: number
+            numGpu: number
           }) => {
             Object.entries(values).forEach(([key, value]) => {
               setModelSetting(key, value)
@@ -106,6 +106,17 @@ export const ModelSettings = () => {
                         placeholder={t("modelSettings.form.topP.placeholder")}
                       />
                     </Form.Item>
+                    <Form.Item
+                      name="numGpu"
+                      label={t("modelSettings.form.numGpu.label")}>
+                      <InputNumber
+                        style={{ width: "100%" }}
+                        size="large"
+                        placeholder={t(
+                          "modelSettings.form.numGpu.placeholder"
+                        )}
+                      />
+                    </Form.Item>
                   </React.Fragment>
                 )
               }

src/components/Option/Settings/rag.tsx

@@ -10,7 +10,7 @@ import {
 } from "~/services/ollama"
 import { SettingPrompt } from "./prompt"
 import { useTranslation } from "react-i18next"
-import { getTotalFilePerKB } from "@/services/app"
+import { getNoOfRetrievedDocs, getTotalFilePerKB } from "@/services/app"
 import { SidepanelRag } from "./sidepanel-rag"
 
 export const RagSettings = () => {
@@ -21,20 +21,28 @@ export const RagSettings = () => {
   const { data: ollamaInfo, status } = useQuery({
     queryKey: ["fetchRAGSettings"],
     queryFn: async () => {
-      const [allModels, chunkOverlap, chunkSize, defaultEM, totalFilePerKB] =
+      const [
-        await Promise.all([
+        allModels,
-          getAllModels({ returnEmpty: true }),
+        chunkOverlap,
-          defaultEmbeddingChunkOverlap(),
+        chunkSize,
-          defaultEmbeddingChunkSize(),
+        defaultEM,
-          defaultEmbeddingModelForRag(),
+        totalFilePerKB,
-          getTotalFilePerKB()
+        noOfRetrievedDocs
-        ])
+      ] = await Promise.all([
+        getAllModels({ returnEmpty: true }),
+        defaultEmbeddingChunkOverlap(),
+        defaultEmbeddingChunkSize(),
+        defaultEmbeddingModelForRag(),
+        getTotalFilePerKB(),
+        getNoOfRetrievedDocs()
+      ])
       return {
         models: allModels,
         chunkOverlap,
         chunkSize,
         defaultEM,
-        totalFilePerKB
+        totalFilePerKB,
+        noOfRetrievedDocs
       }
     }
   })
@@ -45,8 +53,15 @@ export const RagSettings = () => {
       chunkSize: number
       overlap: number
       totalFilePerKB: number
+      noOfRetrievedDocs: number
     }) => {
-      await saveForRag(data.model, data.chunkSize, data.overlap,  data.totalFilePerKB)
+      await saveForRag(
+        data.model,
+        data.chunkSize,
+        data.overlap,
+        data.totalFilePerKB,
+        data.noOfRetrievedDocs
+      )
       return true
     },
     onSuccess: () => {
@@ -75,14 +90,16 @@ export const RagSettings = () => {
                   model: data.defaultEM,
                   chunkSize: data.chunkSize,
                   overlap: data.chunkOverlap,
-                  totalFilePerKB: data.totalFilePerKB
+                  totalFilePerKB: data.totalFilePerKB,
+                  noOfRetrievedDocs: data.noOfRetrievedDocs
                 })
               }}
               initialValues={{
                 chunkSize: ollamaInfo?.chunkSize,
                 chunkOverlap: ollamaInfo?.chunkOverlap,
                 defaultEM: ollamaInfo?.defaultEM,
-                totalFilePerKB: ollamaInfo?.totalFilePerKB
+                totalFilePerKB: ollamaInfo?.totalFilePerKB,
+                noOfRetrievedDocs: ollamaInfo?.noOfRetrievedDocs
               }}>
               <Form.Item
                 name="defaultEM"
@@ -142,6 +159,23 @@ export const RagSettings = () => {
                 />
               </Form.Item>
 
+              <Form.Item
+                name="noOfRetrievedDocs"
+                label={t("rag.ragSettings.noOfRetrievedDocs.label")}
+                rules={[
+                  {
+                    required: true,
+                    message: t("rag.ragSettings.noOfRetrievedDocs.required")
+                  }
+                ]}>
+                <InputNumber
+                  style={{ width: "100%" }}
+                  placeholder={t(
+                    "rag.ragSettings.noOfRetrievedDocs.placeholder"
+                  )}
+                />
+              </Form.Item>
+
               <Form.Item
                 name="totalFilePerKB"
                 label={t("rag.ragSettings.totalFilePerKB.label")}
@@ -153,18 +187,18 @@ export const RagSettings = () => {
                 ]}>
                 <InputNumber
                   style={{ width: "100%" }}
+                  min={1}
                   placeholder={t("rag.ragSettings.totalFilePerKB.placeholder")}
                 />
               </Form.Item>
 
-
               <div className="flex justify-end">
                 <SaveButton disabled={isSaveRAGPending} btnType="submit" />
               </div>
             </Form>
           </div>
 
-          <SidepanelRag  />
+          <SidepanelRag />
 
           <div>
             <div>

src/entries/background.ts

@@ -64,31 +64,31 @@ export default defineBackground({
 
     browser.contextMenus.create({
       id: "summarize-pa",
-      title: "Summarize",
+      title: browser.i18n.getMessage("contextSummarize"),
       contexts: ["selection"]
     })
 
     browser.contextMenus.create({
       id: "explain-pa",
-      title: "Explain",
+      title: browser.i18n.getMessage("contextExplain"),
       contexts: ["selection"]
     })
 
     browser.contextMenus.create({
       id: "rephrase-pa",
-      title: "Rephrase",
+      title: browser.i18n.getMessage("contextRephrase"),
       contexts: ["selection"]
     })
 
     browser.contextMenus.create({
       id: "translate-pg",
-      title: "Translate",
+      title: browser.i18n.getMessage("contextTranslate"),
       contexts: ["selection"]
     })
 
     browser.contextMenus.create({
       id: "custom-pg",
-      title: "Custom",
+      title: browser.i18n.getMessage("contextCustom"),
       contexts: ["selection"]
     })
 

src/entries/options/index.html

@@ -4,6 +4,7 @@
     <title>Page Assist - A Web UI for Local AI Models</title>
     <meta name="viewport" content="width=device-width, initial-scale=1">
     <meta name="manifest.type" content="browser_action" />
+    <meta name="manifest.browser_style" content="false" />
     <link href="~/assets/tailwind.css" rel="stylesheet" />
     <meta charset="utf-8" />
   </head>

src/hooks/useMessage.tsx

@@ -122,7 +122,9 @@ export const useMessage = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -248,7 +250,9 @@ export const useMessage = () => {
           numCtx:
             currentChatModelSettings?.numCtx ??
             userDefaultModelSettings?.numCtx,
-          seed: currentChatModelSettings?.seed
+          seed: currentChatModelSettings?.seed,
+          numGpu:
+            currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
         })
         const response = await questionOllama.invoke(promptForQuestion)
         query = response.content.toString()
@@ -442,7 +446,9 @@ export const useMessage = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -645,7 +651,9 @@ export const useMessage = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -718,7 +726,9 @@ export const useMessage = () => {
           numCtx:
             currentChatModelSettings?.numCtx ??
             userDefaultModelSettings?.numCtx,
-          seed: currentChatModelSettings?.seed
+          seed: currentChatModelSettings?.seed,
+          numGpu:
+            currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
         })
         const response = await questionOllama.invoke(promptForQuestion)
         query = response.content.toString()
@@ -890,7 +900,9 @@ export const useMessage = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -932,8 +944,6 @@ export const useMessage = () => {
     let contentToSave = ""
 
     try {
-
-      
       const prompt = await getPrompt(messageType)
       let humanMessage = new HumanMessage({
         content: [

src/hooks/useMessageOption.tsx

@@ -32,6 +32,7 @@ import { useStorage } from "@plasmohq/storage/hook"
 import { useStoreChatModelSettings } from "@/store/model"
 import { getAllDefaultModelSettings } from "@/services/model-settings"
 import { pageAssistModel } from "@/models"
+import { getNoOfRetrievedDocs } from "@/services/app"
 
 export const useMessageOption = () => {
   const {
@@ -117,7 +118,9 @@ export const useMessageOption = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -190,7 +193,9 @@ export const useMessageOption = () => {
           numCtx:
             currentChatModelSettings?.numCtx ??
             userDefaultModelSettings?.numCtx,
-          seed: currentChatModelSettings?.seed
+          seed: currentChatModelSettings?.seed,
+          numGpu:
+            currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
         })
         const response = await questionOllama.invoke(promptForQuestion)
         query = response.content.toString()
@@ -360,7 +365,9 @@ export const useMessageOption = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -576,7 +583,9 @@ export const useMessageOption = () => {
       topP: currentChatModelSettings?.topP ?? userDefaultModelSettings?.topP,
       numCtx:
         currentChatModelSettings?.numCtx ?? userDefaultModelSettings?.numCtx,
-      seed: currentChatModelSettings?.seed
+      seed: currentChatModelSettings?.seed,
+      numGpu:
+        currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
     })
 
     let newMessage: Message[] = []
@@ -665,13 +674,16 @@ export const useMessageOption = () => {
           numCtx:
             currentChatModelSettings?.numCtx ??
             userDefaultModelSettings?.numCtx,
-          seed: currentChatModelSettings?.seed
+          seed: currentChatModelSettings?.seed,
+          numGpu:
+            currentChatModelSettings?.numGpu ?? userDefaultModelSettings?.numGpu
         })
         const response = await questionOllama.invoke(promptForQuestion)
         query = response.content.toString()
       }
+      const docSize = await getNoOfRetrievedDocs()
 
-      const docs = await vectorstore.similaritySearch(query, 4)
+      const docs = await vectorstore.similaritySearch(query, docSize)
       const context = formatDocs(docs)
       const source = docs.map((doc) => {
         return {

src/models/ChatOllama.ts

@@ -126,7 +126,7 @@ export class ChatOllama
         this.mirostatTau = fields.mirostatTau;
         this.numBatch = fields.numBatch;
         this.numCtx = fields.numCtx;
-        this.numGpu = fields.numGpu;
+        this.numGpu = fields.numGpu === null ? undefined : fields.numGpu;
         this.numGqa = fields.numGqa;
         this.numKeep = fields.numKeep;
         this.numPredict = fields.numPredict;

src/models/index.ts

@@ -9,7 +9,8 @@ export const pageAssistModel = async ({
   topK,
   topP,
   numCtx,
-  seed
+  seed,
+  numGpu
 }: {
   model: string
   baseUrl: string
@@ -19,6 +20,7 @@ export const pageAssistModel = async ({
   topP?: number
   numCtx?: number
   seed?: number
+  numGpu?: number
 }) => {
   switch (model) {
     case "chrome::gemini-nano::page-assist":
@@ -35,7 +37,8 @@ export const pageAssistModel = async ({
         topP,
         numCtx,
         seed,
-        model
+        model,
+        numGpu
       })
   }
 }

src/public/_locales/en/messages.json

@@ -10,5 +10,20 @@
     },
     "openOptionToChat": {
         "message": "Open Web UI to Chat"
+    },
+    "contextSummarize": {
+        "message": "Summarize"
+    },
+    "contextExplain": {
+        "message": "Explain"
+    },
+    "contextRephrase": {
+        "message": "Rephrase"
+    },
+    "contextTranslate" :{
+        "message": "Translate"
+    },
+    "contextCustom": {
+        "message": "Custom"
     }
 }

src/public/_locales/es/messages.json

@@ -7,5 +7,23 @@
     },
     "openSidePanelToChat": {
         "message": "Abrir Copilot para Chatear"
+    },
+    "openOptionToChat": {
+        "message": "Abrir Web UI para Chatear"
+    },
+    "contextSummarize": {
+        "message": "Resumir"
+    },
+    "contextExplain": {
+        "message": "Explicar"
+    },
+    "contextRephrase": {
+        "message": "Reformular"
+    },
+    "contextTranslate": {
+        "message": "Traducir"
+    },
+    "contextCustom": {
+        "message": "Personalizado"
     }
-}
+}

src/public/_locales/fa/messages.json

@@ -10,5 +10,20 @@
     },
     "openOptionToChat": {
         "message": "باز کردن رابط کاربری وب برای گفتگو"
+    },
+    "contextSummarize": {
+        "message": "خلاصه کردن"
+    },
+    "contextExplain": {
+        "message": "توضیح دادن"
+    },
+    "contextRephrase": {
+        "message": "بازنویسی"
+    },
+    "contextTranslate" :{
+        "message": "ترجمه کردن"
+    },
+    "contextCustom": {
+        "message": "سفارشی"
     }
 }

src/public/_locales/fr/messages.json

@@ -7,5 +7,20 @@
     },
     "openSidePanelToChat": {
         "message": "Ouvrir Copilot pour discuter"
+    },
+    "contextSummarize": {
+        "message": "Résumer"
+    },
+    "contextExplain": {
+        "message": "Expliquer"
+    },
+    "contextRephrase": {
+        "message": "Reformuler"
+    },
+    "contextTranslate" :{
+        "message": "Traduire"
+    },
+    "contextCustom": {
+        "message": "Personnalisé"
     }
 }

src/public/_locales/it/messages.json

@@ -7,5 +7,20 @@
     },
     "openSidePanelToChat": {
         "message": "Apri Copilot per chattare"
+    },
+    "contextSummarize": {
+        "message": "Riassumi"
+    },
+    "contextExplain": {
+        "message": "Spiega"
+    },
+    "contextRephrase": {
+        "message": "Riformula"
+    },
+    "contextTranslate" :{
+        "message": "Traduci"
+    },
+    "contextCustom": {
+        "message": "Personalizzato"
     }
 }

src/public/_locales/ja/messages.json

@@ -7,5 +7,20 @@
   },
   "openSidePanelToChat": {
     "message": "チャットするためにCopilotを開く"
-  }
+  },
+  "contextSummarize": {
+        "message": "要約"
+    },
+    "contextExplain": {
+        "message": "説明"
+    },
+    "contextRephrase": {
+        "message": "言い換え"
+    },
+    "contextTranslate" :{
+        "message": "翻訳"
+    },
+    "contextCustom": {
+        "message": "カスタム"
+    }
 }

src/public/_locales/ml/messages.json

@@ -7,5 +7,20 @@
     },
     "openSidePanelToChat": {
         "message": "ചാറ്റ് ചെയ്യാന്‍ സൈഡ് പാനല്‍ തുറക്കുക"
+    },
+    "contextSummarize": {
+        "message": "സംഗ്രഹിക്കുക"
+    },
+    "contextExplain": {
+        "message": "വിശദീകരിക്കുക"
+    },
+    "contextRephrase": {
+        "message": "പുനഃരൂപീകരിക്കുക"
+    },
+    "contextTranslate" :{
+        "message": "വിവർത്തനം ചെയ്യുക"
+    },
+    "contextCustom": {
+        "message": "ഇഷ്ടാനുസൃതം"
     }
 }

src/public/_locales/ru/messages.json

@@ -7,5 +7,20 @@
     },
     "openSidePanelToChat": {
         "message": "Открыть Copilot для чата"
+    },
+    "contextSummarize": {
+        "message": "Обобщить"
+    },
+    "contextExplain": {
+        "message": "Объяснить"
+    },
+    "contextRephrase": {
+        "message": "Перефразировать"
+    },
+    "contextTranslate" :{
+        "message": "Перевести"
+    },
+    "contextCustom": {
+        "message": "Пользовательский"
     }
 }

src/public/_locales/zh_CN/messages.json

@@ -7,5 +7,20 @@
     },
     "openSidePanelToChat": {
         "message": "打开Copilot进行聊天"
+    },
+    "contextSummarize": {
+        "message": "总结"
+    },
+    "contextExplain": {
+        "message": "解释"
+    },
+    "contextRephrase": {
+        "message": "改述"
+    },
+    "contextTranslate" :{
+        "message": "翻译"
+    },
+    "contextCustom": {
+        "message": "自定义"
     }
 }

src/services/app.ts

@@ -108,4 +108,13 @@ export const getTotalFilePerKB = async (): Promise<number> => {
 
 export const setTotalFilePerKB = async (totalFilePerKB: number): Promise<void> => {
   await storage.set("totalFilePerKB", totalFilePerKB);
-};
+};
+
+export const getNoOfRetrievedDocs = async (): Promise<number> => {
+  const noOfRetrievedDocs = await storage.get<number>("noOfRetrievedDocs");
+  return noOfRetrievedDocs || 4 
+}
+
+export const setNoOfRetrievedDocs = async (noOfRetrievedDocs: number): Promise<void> => {
+  await storage.set("noOfRetrievedDocs", noOfRetrievedDocs);
+}

src/services/ollama.ts

@@ -2,7 +2,7 @@ import { Storage } from "@plasmohq/storage"
 import { cleanUrl } from "../libs/clean-url"
 import { urlRewriteRuntime } from "../libs/runtime"
 import { getChromeAIModel } from "./chrome"
-import { setTotalFilePerKB } from "./app"
+import { setNoOfRetrievedDocs, setTotalFilePerKB } from "./app"
 
 const storage = new Storage()
 
@@ -326,12 +326,14 @@ export const saveForRag = async (
   model: string,
   chunkSize: number,
   overlap: number,
-  totalFilePerKB: number
+  totalFilePerKB: number,
+  noOfRetrievedDocs: number
 ) => {
   await setDefaultEmbeddingModelForRag(model)
   await setDefaultEmbeddingChunkSize(chunkSize)
   await setDefaultEmbeddingChunkOverlap(overlap)
   await setTotalFilePerKB(totalFilePerKB)
+  await setNoOfRetrievedDocs(noOfRetrievedDocs)
 }
 
 export const getWebSearchPrompt = async () => {

wxt.config.ts

@@ -37,7 +37,7 @@ export default defineConfig({
       topLevelAwait({
         promiseExportName: "__tla",
         promiseImportName: (i) => `__tla_${i}`
-      })
+      }) as any
     ],
     build: {
       rollupOptions: {
@@ -50,7 +50,7 @@ export default defineConfig({
   outDir: "build",
 
   manifest: {
-    version: "1.2.0",
+    version: "1.2.1",
     name:
       process.env.TARGET === "firefox"
         ? "Page Assist - A Web UI for Local AI Models"
@@ -90,4 +90,4 @@ export default defineConfig({
         ? firefoxMV2Permissions
         : chromeMV3Permissions
   }
-})
+}) as any