Merge pull request #2785 from cheahjs/feat/openai-embeddings-batch

feat: add RAG_EMBEDDING_OPENAI_BATCH_SIZE to batch multiple embeddings

src/lib/apis/rag/index.ts

@@ -415,6 +415,7 @@ export const getEmbeddingConfig = async (token: string) => {
 type OpenAIConfigForm = {
 	key: string;
 	url: string;
+	batch_size: number;
 };
 
 type EmbeddingModelUpdateForm = {

src/lib/components/documents/Settings/General.svelte

@@ -31,6 +31,7 @@
 
 	let OpenAIKey = '';
 	let OpenAIUrl = '';
+	let OpenAIBatchSize = 1;
 
 	let querySettings = {
 		template: '',
@@ -92,7 +93,8 @@
 				? {
 						openai_config: {
 							key: OpenAIKey,
-							url: OpenAIUrl
+							url: OpenAIUrl,
+							batch_size: OpenAIBatchSize
 						}
 				  }
 				: {})
@@ -159,6 +161,7 @@
 
 			OpenAIKey = embeddingConfig.openai_config.key;
 			OpenAIUrl = embeddingConfig.openai_config.url;
+			OpenAIBatchSize = embeddingConfig.openai_config.batch_size ?? 1;
 		}
 	};
 
@@ -282,6 +285,30 @@
 						required
 					/>
 				</div>
+				<div class="flex mt-0.5 space-x-2">
+					<div class=" self-center text-xs font-medium">{$i18n.t('Embedding Batch Size')}</div>
+					<div class=" flex-1">
+						<input
+							id="steps-range"
+							type="range"
+							min="1"
+							max="2048"
+							step="1"
+							bind:value={OpenAIBatchSize}
+							class="w-full h-2 rounded-lg appearance-none cursor-pointer dark:bg-gray-700"
+						/>
+					</div>
+					<div class="">
+						<input
+							bind:value={OpenAIBatchSize}
+							type="number"
+							class=" bg-transparent text-center w-14"
+							min="-2"
+							max="16000"
+							step="1"
+						/>
+					</div>
+				</div>
 			{/if}
 
 			<div class=" flex w-full justify-between">

src/lib/i18n/locales/ar-BH/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "تعديل الملف",
 	"Edit User": "تعديل المستخدم",
 	"Email": "البريد",
+	"Embedding Batch Size": "",
 	"Embedding Model": "نموذج التضمين",
 	"Embedding Model Engine": "تضمين محرك النموذج",
 	"Embedding model set to \"{{embedding_model}}\"": "تم تعيين نموذج التضمين على \"{{embedding_model}}\"",

src/lib/i18n/locales/bg-BG/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Редактиране на документ",
 	"Edit User": "Редактиране на потребител",
 	"Email": "Имейл",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Модел за вграждане",
 	"Embedding Model Engine": "Модел за вграждане",
 	"Embedding model set to \"{{embedding_model}}\"": "Модел за вграждане е настроен на \"{{embedding_model}}\"",

src/lib/i18n/locales/bn-BD/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "ডকুমেন্ট এডিট করুন",
 	"Edit User": "ইউজার এডিট করুন",
 	"Email": "ইমেইল",
+	"Embedding Batch Size": "",
 	"Embedding Model": "ইমেজ ইমেবডিং মডেল",
 	"Embedding Model Engine": "ইমেজ ইমেবডিং মডেল ইঞ্জিন",
 	"Embedding model set to \"{{embedding_model}}\"": "ইমেজ ইমেবডিং মডেল সেট করা হয়েছে - \"{{embedding_model}}\"",

src/lib/i18n/locales/ca-ES/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Edita Document",
 	"Edit User": "Edita Usuari",
 	"Email": "Correu electrònic",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Model d'embutiment",
 	"Embedding Model Engine": "Motor de model d'embutiment",
 	"Embedding model set to \"{{embedding_model}}\"": "Model d'embutiment configurat a \"{{embedding_model}}\"",

src/lib/i18n/locales/ceb-PH/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "I-edit ang dokumento",
 	"Edit User": "I-edit ang tiggamit",
 	"Email": "E-mail",
+	"Embedding Batch Size": "",
 	"Embedding Model": "",
 	"Embedding Model Engine": "",
 	"Embedding model set to \"{{embedding_model}}\"": "",

src/lib/i18n/locales/de-DE/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Dokument bearbeiten",
 	"Edit User": "Benutzer bearbeiten",
 	"Email": "E-Mail",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Embedding-Modell",
 	"Embedding Model Engine": "Embedding-Modell-Engine",
 	"Embedding model set to \"{{embedding_model}}\"": "Embedding-Modell auf \"{{embedding_model}}\" gesetzt",

src/lib/i18n/locales/dg-DG/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Edit Doge",
 	"Edit User": "Edit Wowser",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "",
 	"Embedding Model Engine": "",
 	"Embedding model set to \"{{embedding_model}}\"": "",

src/lib/i18n/locales/en-GB/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "",
 	"Edit User": "",
 	"Email": "",
+	"Embedding Batch Size": "",
 	"Embedding Model": "",
 	"Embedding Model Engine": "",
 	"Embedding model set to \"{{embedding_model}}\"": "",

src/lib/i18n/locales/en-US/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "",
 	"Edit User": "",
 	"Email": "",
+	"Embedding Batch Size": "",
 	"Embedding Model": "",
 	"Embedding Model Engine": "",
 	"Embedding model set to \"{{embedding_model}}\"": "",

src/lib/i18n/locales/es-ES/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Editar Documento",
 	"Edit User": "Editar Usuario",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modelo de Embedding",
 	"Embedding Model Engine": "Motor de Modelo de Embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modelo de Embedding configurado a \"{{embedding_model}}\"",

src/lib/i18n/locales/fa-IR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "ویرایش سند",
 	"Edit User": "ویرایش کاربر",
 	"Email": "ایمیل",
+	"Embedding Batch Size": "",
 	"Embedding Model": "مدل پیدائش",
 	"Embedding Model Engine": "محرک مدل پیدائش",
 	"Embedding model set to \"{{embedding_model}}\"": "مدل پیدائش را به \"{{embedding_model}}\" تنظیم کنید",

src/lib/i18n/locales/fi-FI/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Muokkaa asiakirjaa",
 	"Edit User": "Muokkaa käyttäjää",
 	"Email": "Sähköposti",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Upotusmalli",
 	"Embedding Model Engine": "Upotusmallin moottori",
 	"Embedding model set to \"{{embedding_model}}\"": "\"{{embedding_model}}\" valittu upotusmalliksi",

src/lib/i18n/locales/fr-CA/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Éditer le document",
 	"Edit User": "Éditer l'utilisateur",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modèle d'embedding",
 	"Embedding Model Engine": "Moteur du modèle d'embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modèle d'embedding défini sur \"{{embedding_model}}\"",

src/lib/i18n/locales/fr-FR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Éditer le document",
 	"Edit User": "Éditer l'utilisateur",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modèle pour l'Embedding",
 	"Embedding Model Engine": "Moteur du Modèle d'Embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modèle d'embedding défini sur \"{{embedding_model}}\"",

src/lib/i18n/locales/he-IL/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "ערוך מסמך",
 	"Edit User": "ערוך משתמש",
 	"Email": "דוא\"ל",
+	"Embedding Batch Size": "",
 	"Embedding Model": "מודל הטמעה",
 	"Embedding Model Engine": "מנוע מודל הטמעה",
 	"Embedding model set to \"{{embedding_model}}\"": "מודל ההטמעה הוגדר ל-\"{{embedding_model}}\"",

src/lib/i18n/locales/hi-IN/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "दस्तावेज़ संपादित करें",
 	"Edit User": "यूजर को संपादित करो",
 	"Email": "ईमेल",
+	"Embedding Batch Size": "",
 	"Embedding Model": "मॉडेल अनुकूलन",
 	"Embedding Model Engine": "एंबेडिंग मॉडल इंजन",
 	"Embedding model set to \"{{embedding_model}}\"": "एम्बेडिंग मॉडल को \"{{embedding_model}}\" पर सेट किया गया",

src/lib/i18n/locales/hr-HR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Uredi dokument",
 	"Edit User": "Uredi korisnika",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Embedding model",
 	"Embedding Model Engine": "Embedding model pogon",
 	"Embedding model set to \"{{embedding_model}}\"": "Embedding model postavljen na \"{{embedding_model}}\"",

src/lib/i18n/locales/it-IT/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Modifica documento",
 	"Edit User": "Modifica utente",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modello di embedding",
 	"Embedding Model Engine": "Motore del modello di embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modello di embedding impostato su \"{{embedding_model}}\"",

src/lib/i18n/locales/ja-JP/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "ドキュメントを編集",
 	"Edit User": "ユーザーを編集",
 	"Email": "メールアドレス",
+	"Embedding Batch Size": "",
 	"Embedding Model": "埋め込みモデル",
 	"Embedding Model Engine": "埋め込みモデルエンジン",
 	"Embedding model set to \"{{embedding_model}}\"": "埋め込みモデルを\"{{embedding_model}}\"に設定しました",

src/lib/i18n/locales/ka-GE/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "დოკუმენტის ედიტირება",
 	"Edit User": "მომხმარებლის ედიტირება",
 	"Email": "ელ-ფოსტა",
+	"Embedding Batch Size": "",
 	"Embedding Model": "ჩასმის ძირითადი პროგრამა",
 	"Embedding Model Engine": "ჩასმის ძირითადი პროგრამა",
 	"Embedding model set to \"{{embedding_model}}\"": "ჩასმის ძირითადი პროგრამა ჩართულია \"{{embedding_model}}\"",

src/lib/i18n/locales/ko-KR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "문서 편집",
 	"Edit User": "사용자 편집",
 	"Email": "이메일",
+	"Embedding Batch Size": "",
 	"Embedding Model": "임베딩 모델",
 	"Embedding Model Engine": "임베딩 모델 엔진",
 	"Embedding model set to \"{{embedding_model}}\"": "임베딩 모델을 \"{{embedding_model}}\"로 설정됨",

src/lib/i18n/locales/nl-NL/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Wijzig Doc",
 	"Edit User": "Wijzig Gebruiker",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Embedding Model",
 	"Embedding Model Engine": "Embedding Model Engine",
 	"Embedding model set to \"{{embedding_model}}\"": "Embedding model ingesteld op \"{{embedding_model}}\"",

src/lib/i18n/locales/pa-IN/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "ਡਾਕੂਮੈਂਟ ਸੰਪਾਦਨ ਕਰੋ",
 	"Edit User": "ਉਪਭੋਗਤਾ ਸੰਪਾਦਨ ਕਰੋ",
 	"Email": "ਈਮੇਲ",
+	"Embedding Batch Size": "",
 	"Embedding Model": "ਐਮਬੈੱਡਿੰਗ ਮਾਡਲ",
 	"Embedding Model Engine": "ਐਮਬੈੱਡਿੰਗ ਮਾਡਲ ਇੰਜਣ",
 	"Embedding model set to \"{{embedding_model}}\"": "ਐਮਬੈੱਡਿੰਗ ਮਾਡਲ ਨੂੰ \"{{embedding_model}}\" 'ਤੇ ਸੈੱਟ ਕੀਤਾ ਗਿਆ",

src/lib/i18n/locales/pl-PL/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Edytuj dokument",
 	"Edit User": "Edytuj użytkownika",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Model osadzania",
 	"Embedding Model Engine": "Silnik modelu osadzania",
 	"Embedding model set to \"{{embedding_model}}\"": "Model osadzania ustawiono na \"{{embedding_model}}\"",

src/lib/i18n/locales/pt-BR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Editar Documento",
 	"Edit User": "Editar Usuário",
 	"Email": "E-mail",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modelo de Embedding",
 	"Embedding Model Engine": "Motor de Modelo de Embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modelo de Embedding definido como \"{{embedding_model}}\"",

src/lib/i18n/locales/pt-PT/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Editar Documento",
 	"Edit User": "Editar Usuário",
 	"Email": "E-mail",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Modelo de Embedding",
 	"Embedding Model Engine": "Motor de Modelo de Embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Modelo de Embedding definido como \"{{embedding_model}}\"",

src/lib/i18n/locales/ru-RU/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Редактировать документ",
 	"Edit User": "Редактировать пользователя",
 	"Email": "Электронная почта",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Модель эмбеддинга",
 	"Embedding Model Engine": "Модель эмбеддинга",
 	"Embedding model set to \"{{embedding_model}}\"": "Эмбеддинг-модель установлена в \"{{embedding_model}}\"",

src/lib/i18n/locales/sr-RS/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Уреди документ",
 	"Edit User": "Уреди корисника",
 	"Email": "Е-пошта",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Модел уградње",
 	"Embedding Model Engine": "Мотор модела уградње",
 	"Embedding model set to \"{{embedding_model}}\"": "Модел уградње подешен на \"{{embedding_model}}\"",

src/lib/i18n/locales/sv-SE/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Redigera dokument",
 	"Edit User": "Redigera användare",
 	"Email": "E-post",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Embeddingsmodell",
 	"Embedding Model Engine": "Embeddingsmodellmotor",
 	"Embedding model set to \"{{embedding_model}}\"": "Embeddingsmodell inställd på \"{{embedding_model}}\"",

src/lib/i18n/locales/tr-TR/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Belgeyi Düzenle",
 	"Edit User": "Kullanıcıyı Düzenle",
 	"Email": "E-posta",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Gömme Modeli",
 	"Embedding Model Engine": "Gömme Modeli Motoru",
 	"Embedding model set to \"{{embedding_model}}\"": "Gömme modeli \"{{embedding_model}}\" olarak ayarlandı",

src/lib/i18n/locales/uk-UA/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Редагувати документ",
 	"Edit User": "Редагувати користувача",
 	"Email": "Електронна пошта",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Модель вбудовування",
 	"Embedding Model Engine": "Двигун модели встраивания ",
 	"Embedding model set to \"{{embedding_model}}\"": "Встановлена модель вбудовування \"{{embedding_model}}\"",

src/lib/i18n/locales/vi-VN/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "Thay đổi tài liệu",
 	"Edit User": "Thay đổi thông tin người sử dụng",
 	"Email": "Email",
+	"Embedding Batch Size": "",
 	"Embedding Model": "Mô hình embedding",
 	"Embedding Model Engine": "Trình xử lý embedding",
 	"Embedding model set to \"{{embedding_model}}\"": "Mô hình embedding đã được thiết lập thành \"{{embedding_model}}\"",

src/lib/i18n/locales/zh-CN/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "编辑文档",
 	"Edit User": "编辑用户",
 	"Email": "邮箱地址",
+  "Embedding Batch Size": "",
 	"Embedding Model": "语义向量模型",
 	"Embedding Model Engine": "语义向量模型引擎",
 	"Embedding model set to \"{{embedding_model}}\"": "语义向量模型设置为 \"{{embedding_model}}\"",

src/lib/i18n/locales/zh-TW/translation.json

@@ -170,6 +170,7 @@
 	"Edit Doc": "編輯文件",
 	"Edit User": "編輯使用者",
 	"Email": "電子郵件",
+	"Embedding Batch Size": "",
 	"Embedding Model": "嵌入模型",
 	"Embedding Model Engine": "嵌入模型引擎",
 	"Embedding model set to \"{{embedding_model}}\"": "嵌入模型已設定為 \"{{embedding_model}}\"",