Merge pull request #709 from janhq/fix/extends-from-open-ai

chore: update models to extends from openai

Author: namchuai

cortex-js/src/app.module.ts

@@ -10,7 +10,6 @@ import { ExtensionModule } from './infrastructure/repositories/extensions/extens
 import { CortexModule } from './usecases/cortex/cortex.module';
 import { ConfigModule } from '@nestjs/config';
 import { env } from 'node:process';
-import { SeedService } from './usecases/seed/seed.service';
 import { FileManagerModule } from './infrastructure/services/file-manager/file-manager.module';
 import { AppLoggerMiddleware } from './infrastructure/middlewares/app.logger.middleware';
 import { TelemetryModule } from './usecases/telemetry/telemetry.module';
@@ -38,6 +37,7 @@ import { DownloadManagerModule } from './infrastructure/services/download-manage
       envFilePath: env.NODE_ENV !== 'production' ? '.env.development' : '.env',
     }),
     EventEmitterModule.forRoot(),
+    DownloadManagerModule,
     DatabaseModule,
     MessagesModule,
     ThreadsModule,
@@ -62,7 +62,6 @@ import { DownloadManagerModule } from './infrastructure/services/download-manage
     EventsController,
   ],
   providers: [
-    SeedService,
     {
       provide: APP_FILTER,
       useClass: GlobalExceptionFilter,

cortex-js/src/domain/models/assistant.interface.ts

@@ -1,39 +1,10 @@
-/**
- * Assistant type defines the shape of an assistant object.
- * @stored
- */
+import { Assistant as OpenAiAssistant } from 'openai/resources/beta/assistants';
+import { AssistantResponseFormatOption as OpenAIAssistantResponseFormatOption } from 'openai/resources/beta/threads/threads';
 
-export interface AssistantTool {
-  type: string;
-  enabled: boolean;
-  settings: any;
+export interface Assistant extends OpenAiAssistant {
+  avatar?: string;
 }
 
-export interface Assistant {
-  /** Represents the unique identifier of the object. */
-  id: string;
-  /** Represents the avatar of the user. */
-  avatar: string;
-  /** Represents the location of the thread. */
-  thread_location?: string;
-  /** Represents the object. */
-  object: string;
-  /** Represents the creation timestamp of the object. */
-  created_at: number;
-  /** Represents the name of the object. */
-  name: string;
-  /** Represents the description of the object. */
-  description?: string;
-  /** Represents the model of the object. */
-  model: string;
-  /** Represents the instructions for the object. */
-  instructions?: string;
-  /** Represents the tools associated with the object. */
-  tools?: AssistantTool[];
-  /** Represents the file identifiers associated with the object. */
-  file_ids: string[];
-  /** Represents the metadata of the object. */
-  metadata?: AssistantMetadata;
-}
+export type AssistantResponseFormatOption = OpenAIAssistantResponseFormatOption;
 
-export interface AssistantMetadata {}
+export interface AssistantToolResources extends OpenAiAssistant.ToolResources {}

cortex-js/src/domain/models/message.interface.ts

@@ -1,81 +1,16 @@
-export enum ChatCompletionRole {
-  System = 'system',
-  Assistant = 'assistant',
-  User = 'user',
-}
+import {
+  Message as OpenAiMessage,
+  MessageContent as OpenAiMessageContent,
+  TextContentBlock as OpenAiTextContentBlock,
+} from 'openai/resources/beta/threads/messages';
 
-export enum ContentType {
-  Text = 'text',
-  Image = 'image',
-  Pdf = 'pdf',
-}
+export interface Message extends OpenAiMessage {}
 
-export interface ContentValue {
-  value: string;
-  annotations: string[];
-  name?: string;
-  size?: number;
-}
+export type MessageContent = OpenAiMessageContent;
 
-export interface ThreadContent {
-  type: ContentType;
-  text: ContentValue;
-}
+export type TextContentBlock = OpenAiTextContentBlock;
 
-/**
- * The status of the message.
- * @data_transfer_object
- */
-export enum MessageStatus {
-  /** Message is fully loaded. **/
-  Ready = 'ready',
-  /** Message is not fully loaded. **/
-  Pending = 'pending',
-  /** Message loaded with error. **/
-  Error = 'error',
-  /** Message is cancelled streaming */
-  Stopped = 'stopped',
-}
+export interface MessageIncompleteDetails
+  extends OpenAiMessage.IncompleteDetails {}
 
-/**
- * The error code which explain what error type. Used in conjunction with MessageStatus.Error
- */
-export enum ErrorCode {
-  InvalidApiKey = 'invalid_api_key',
-
-  InsufficientQuota = 'insufficient_quota',
-
-  InvalidRequestError = 'invalid_request_error',
-
-  Unknown = 'unknown',
-}
-
-export interface Message {
-  /** Unique identifier for the message, generated by default using the ULID method. **/
-  id: string;
-  /** Object name **/
-  object: string;
-  /** Thread id, default is a ulid. **/
-  thread_id: string;
-  /** The assistant id of this thread. **/
-  assistant_id?: string;
-  /** The role of the author of this message. **/
-  role: ChatCompletionRole;
-  /** The content of this message. **/
-  content: ThreadContent[];
-  /** The status of this message. **/
-  status: MessageStatus;
-  /** The timestamp indicating when this message was created. Represented in Unix time. **/
-  created: number;
-  /** The timestamp indicating when this message was updated. Represented in Unix time. **/
-  updated?: number;
-  /** The additional metadata of this message. **/
-  metadata?: MessageMetadata;
-  /** The error code which explain what error type. Used in conjunction with MessageStatus.Error */
-  error_code?: ErrorCode;
-}
-
-/**
- * The additional metadata of this message.
- */
-export interface MessageMetadata {}
+export interface MessageAttachment extends OpenAiMessage.Attachment {}

cortex-js/src/domain/models/model.interface.ts

@@ -1,8 +1,9 @@
-/**
- * Model type defines the shape of a model object.
- * @stored
- */
-export interface Model {
+import { Model as OpenAiModel } from 'openai/resources/models';
+
+export interface Model
+  extends OpenAiModel,
+    ModelSettingParams,
+    ModelRuntimeParams {
   /**
    * Model identifier.
    */
@@ -23,57 +24,23 @@ export interface Model {
    */
   files: string[] | ModelArtifact;
 
-  /**
-   * GGUF metadata: tokenizer.chat_template
-   */
-  prompt_template?: string;
-
-  /**
-   * Defines specific tokens or phrases at which the model will stop generating further output.
-   */
-  stop?: string[];
-
-  /// Inferencing
-  /**
-   * Set probability threshold for more relevant outputs.
-   */
-  top_p?: number;
-
-  /**
-   * Controls the randomness of the model’s output.
-   */
-  temperature?: number;
-
-  /**
-   * Adjusts the likelihood of the model repeating words or phrases in its output.
-   */
-  frequency_penalty?: number;
-
-  /**
-   * Influences the generation of new and varied concepts in the model’s output.
-   */
-  presence_penalty?: number;
+  metadata?: Record<string, any>;
+}
 
-  /// Engines
+/**
+ * The available model settings.
+ */
+export interface ModelSettingParams {
   /**
    * The context length for model operations varies; the maximum depends on the specific model used.
    */
   ctx_len?: number;
 
-  /**
-   * Enable real-time data processing for faster predictions.
-   */
-  stream?: boolean;
-
-  /*
-   * The maximum number of tokens the model will generate in a single response.
-   */
-  max_tokens?: number;
-
   /**
    * The number of layers to load onto the GPU for acceleration.
    */
   ngl?: number;
+  embedding?: boolean;
 
   /**
    * Number of parallel sequences to decode
@@ -85,6 +52,22 @@ export interface Model {
    */
   cpu_threads?: number;
 
+  /**
+   * GGUF metadata: tokenizer.chat_template
+   */
+  prompt_template?: string;
+  system_prompt?: string;
+  ai_prompt?: string;
+  user_prompt?: string;
+  llama_model_path?: string;
+  mmproj?: string;
+  cont_batching?: boolean;
+
+  /**
+   * The model engine.
+   */
+  engine?: string;
+
   /**
    * The prompt to use for internal configuration
    */
@@ -134,59 +117,48 @@ export interface Model {
    * To enable mmap, default is true
    */
   use_mmap?: boolean;
-
-  /**
-   * The model engine.
-   */
-  engine?: string;
-}
-
-/**
- * The available model settings.
- */
-export interface ModelSettingParams {
-  ctx_len?: number;
-  ngl?: number;
-  embedding?: boolean;
-  n_parallel?: number;
-  cpu_threads?: number;
-  prompt_template?: string;
-  system_prompt?: string;
-  ai_prompt?: string;
-  user_prompt?: string;
-  llama_model_path?: string;
-  mmproj?: string;
-  cont_batching?: boolean;
-  engine?: string;
-  stop?: string[];
-  pre_prompt?: string;
-  n_batch?: number;
-  caching_enabled?: boolean;
-  grp_attn_n?: number;
-  grp_attn_w?: number;
-  mlock?: boolean;
-  grammar_file?: string;
-  model_type?: string;
-  model_alias?: string;
-  flash_attn?: boolean;
-  cache_type?: string;
-  use_mmap?: boolean;
 }
 
 /**
  * The available model runtime parameters.
  */
 export interface ModelRuntimeParams {
+  /**
+   * Controls the randomness of the model’s output.
+   */
   temperature?: number;
   token_limit?: number;
   top_k?: number;
+
+  /**
+   * Set probability threshold for more relevant outputs.
+   */
   top_p?: number;
+
+  /**
+   * Enable real-time data processing for faster predictions.
+   */
   stream?: boolean;
+
+  /*
+   * The maximum number of tokens the model will generate in a single response.
+   */
   max_tokens?: number;
+
+  /**
+   * Defines specific tokens or phrases at which the model will stop generating further output.
+   */
   stop?: string[];
+
+  /**
+   * Adjusts the likelihood of the model repeating words or phrases in its output.
+   */
   frequency_penalty?: number;
+
+  /**
+   * Influences the generation of new and varied concepts in the model’s output.
+   */
   presence_penalty?: number;
-  engine?: string;
 }
 
 /**