dashscope 1.8.0__py3-none-any.whl → 1.25.6__py3-none-any.whl

dashscope/embeddings/multimodal_embedding.py

@@ -1,11 +1,14 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 from dataclasses import dataclass
 from typing import List
 
-from dashscope.api_entities.dashscope_response import (DashScopeAPIResponse, 
+from dashscope.api_entities.dashscope_response import (DashScopeAPIResponse,
                                                        DictMixin)
-from dashscope.client.base_api import BaseApi
+from dashscope.client.base_api import BaseApi, BaseAioApi
 from dashscope.common.error import InputRequired, ModelRequired
 from dashscope.common.utils import _get_task_group_and_task
+from dashscope.utils.oss_utils import preprocess_message_element
 
 
 @dataclass(init=False)
@@ -15,30 +18,34 @@ class MultiModalEmbeddingItemBase(DictMixin):
     def __init__(self, factor: float, **kwargs):
         super().__init__(factor=factor, **kwargs)
 
+
 @dataclass(init=False)
 class MultiModalEmbeddingItemText(MultiModalEmbeddingItemBase):
     text: str
-    
-    def __init__(self, text: str,  factor: float, **kwargs):
+
+    def __init__(self, text: str, factor: float, **kwargs):
         super().__init__(factor, **kwargs)
         self.text = text
 
+
 @dataclass(init=False)
 class MultiModalEmbeddingItemImage(MultiModalEmbeddingItemBase):
     image: str
-    
+
     def __init__(self, image: str, factor: float, **kwargs):
         super().__init__(factor, **kwargs)
         self.image = image
-        
+
+
 @dataclass(init=False)
 class MultiModalEmbeddingItemAudio(MultiModalEmbeddingItemBase):
     audio: str
-    
+
     def __init__(self, audio: str, factor: float, **kwargs):
         super().__init__(factor, **kwargs)
         self.audio = audio
-        
+
+
 class MultiModalEmbedding(BaseApi):
     task = 'multimodal-embedding'
 
@@ -46,7 +53,11 @@ class MultiModalEmbedding(BaseApi):
         multimodal_embedding_one_peace_v1 = 'multimodal-embedding-one-peace-v1'
 
     @classmethod
-    def call(cls, model: str, input: List[MultiModalEmbeddingItemBase],
+    def call(cls,
+             model: str,
+             input: List[MultiModalEmbeddingItemBase],
+             api_key: str = None,
+             workspace: str = None,
              **kwargs) -> DashScopeAPIResponse:
         """Get embedding multimodal contents..
 
@@ -54,8 +65,9 @@ class MultiModalEmbedding(BaseApi):
             model (str): The embedding model name.
             input (List[MultiModalEmbeddingElement]): The embedding elements,
                 every element include data, modal, factor field.
+            workspace (str): The dashscope workspace id.
             **kwargs:
-                auto_truncation(bool, `optional`): Automatically truncate 
+                auto_truncation(bool, `optional`): Automatically truncate
                 audio longer than 15 seconds or text longer than 70 words.
                 Default to false(Too long input will result in failure).
 
@@ -67,6 +79,11 @@ class MultiModalEmbedding(BaseApi):
         if model is None or not model:
             raise ModelRequired('Model is required!')
         embedding_input = {}
+        has_upload = cls._preprocess_message_inputs(model, input, api_key)
+        if has_upload:
+            headers = kwargs.pop('headers', {})
+            headers['X-DashScope-OssResourceResolve'] = 'enable'
+            kwargs['headers'] = headers
         embedding_input['contents'] = input
         kwargs.pop('stream', False)  # not support streaming output.
         task_group, function = _get_task_group_and_task(__name__)
@@ -75,4 +92,95 @@ class MultiModalEmbedding(BaseApi):
                             task_group=task_group,
                             task=MultiModalEmbedding.task,
                             function=function,
+                            api_key=api_key,
+                            workspace=workspace,
                             **kwargs)
+
+    @classmethod
+    def _preprocess_message_inputs(cls, model: str, input: List[dict],
+                                   api_key: str):
+        """preprocess following inputs
+        input = [{'factor': 1, 'text': 'hello'},
+                {'factor': 2, 'audio': ''},
+                {'factor': 3, 'image': ''}]
+        """
+        has_upload = False
+        upload_certificate = None
+        for elem in input:
+            if not isinstance(elem, (int, float, bool, str, bytes, bytearray)):
+                is_upload, upload_certificate = preprocess_message_element(
+                    model, elem, api_key, upload_certificate)
+                if is_upload and not has_upload:
+                    has_upload = True
+        return has_upload
+
+
+class AioMultiModalEmbedding(BaseAioApi):
+    task = 'multimodal-embedding'
+
+    class Models:
+        multimodal_embedding_one_peace_v1 = 'multimodal-embedding-one-peace-v1'
+
+    @classmethod
+    async def call(cls,
+         model: str,
+         input: List[MultiModalEmbeddingItemBase],
+         api_key: str = None,
+         workspace: str = None,
+         **kwargs) -> DashScopeAPIResponse:
+        """Get embedding multimodal contents..
+
+        Args:
+            model (str): The embedding model name.
+            input (List[MultiModalEmbeddingElement]): The embedding elements,
+                every element include data, modal, factor field.
+            workspace (str): The dashscope workspace id.
+            **kwargs:
+                auto_truncation(bool, `optional`): Automatically truncate
+                audio longer than 15 seconds or text longer than 70 words.
+                Default to false(Too long input will result in failure).
+
+        Returns:
+            DashScopeAPIResponse: The embedding result.
+        """
+        if input is None or not input:
+            raise InputRequired('prompt is required!')
+        if model is None or not model:
+            raise ModelRequired('Model is required!')
+        embedding_input = {}
+        has_upload = cls._preprocess_message_inputs(model, input, api_key)
+        if has_upload:
+            headers = kwargs.pop('headers', {})
+            headers['X-DashScope-OssResourceResolve'] = 'enable'
+            kwargs['headers'] = headers
+        embedding_input['contents'] = input
+        kwargs.pop('stream', False)  # not support streaming output.
+        task_group, function = _get_task_group_and_task(__name__)
+        response = await super().call(
+            model=model,
+            input=embedding_input,
+            task_group=task_group,
+            task=MultiModalEmbedding.task,
+            function=function,
+            api_key=api_key,
+            workspace=workspace,
+            **kwargs)
+        return response
+
+    @classmethod
+    def _preprocess_message_inputs(cls, model: str, input: List[dict],
+                                   api_key: str):
+        """preprocess following inputs
+        input = [{'factor': 1, 'text': 'hello'},
+                {'factor': 2, 'audio': ''},
+                {'factor': 3, 'image': ''}]
+        """
+        has_upload = False
+        upload_certificate = None
+        for elem in input:
+            if not isinstance(elem, (int, float, bool, str, bytes, bytearray)):
+                is_upload, upload_certificate = preprocess_message_element(
+                    model, elem, api_key, upload_certificate)
+                if is_upload and not has_upload:
+                    has_upload = True
+        return has_upload

dashscope/embeddings/text_embedding.py

@@ -1,3 +1,5 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 from typing import List, Union
 
 from dashscope.api_entities.dashscope_response import DashScopeAPIResponse
@@ -11,9 +13,16 @@ class TextEmbedding(BaseApi):
 
     class Models:
         text_embedding_v1 = 'text-embedding-v1'
+        text_embedding_v2 = 'text-embedding-v2'
+        text_embedding_v3 = 'text-embedding-v3'
+        text_embedding_v4 = 'text-embedding-v4'
 
     @classmethod
-    def call(cls, model: str, input: Union[str, List[str]],
+    def call(cls,
+             model: str,
+             input: Union[str, List[str]],
+             workspace: str = None,
+             api_key: str = None,
              **kwargs) -> DashScopeAPIResponse:
         """Get embedding of text input.
 
@@ -23,6 +32,7 @@ class TextEmbedding(BaseApi):
                 can be a text or list of text or opened file object,
                 if opened file object, will read all lines,
                 one embedding per line.
+            workspace (str): The dashscope workspace id.
             **kwargs:
                 text_type(str, `optional`): query or document.
 
@@ -41,4 +51,6 @@ class TextEmbedding(BaseApi):
                             task_group=task_group,
                             task=TextEmbedding.task,
                             function=function,
+                            api_key=api_key,
+                            workspace=workspace,
                             **kwargs)

dashscope/{file.py → files.py}

@@ -1,3 +1,5 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 import os
 
 from dashscope.api_entities.dashscope_response import DashScopeAPIResponse
@@ -8,7 +10,7 @@ from dashscope.common.error import InvalidFileFormat
 from dashscope.common.utils import is_validate_fine_tune_file
 
 
-class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
+class Files(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
     SUB_PATH = 'files'
 
     @classmethod
@@ -17,6 +19,7 @@ class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
                purpose: str = FilePurpose.fine_tune,
                description: str = None,
                api_key: str = None,
+               workspace: str = None,
                **kwargs) -> DashScopeAPIResponse:
         """Upload file for model fine-tune or other tasks.
 
@@ -25,6 +28,7 @@ class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
             purpose (str): The purpose of the file[fine-tune|inference]
             description (str, optional): The file description message.
             api_key (str, optional): The api key. Defaults to None.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: The upload information
@@ -39,6 +43,7 @@ class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
                                   descriptions=[description]
                                   if description is not None else None,
                                   api_key=api_key,
+                                  workspace=workspace,
                                   **kwargs)
 
     @classmethod
@@ -46,6 +51,7 @@ class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
              page=1,
              page_size=10,
              api_key: str = None,
+             workspace: str = None,
              **kwargs) -> DashScopeAPIResponse:
         """List uploaded files.
 
@@ -55,40 +61,49 @@ class File(FileUploadMixin, ListMixin, DeleteMixin, GetMixin):
                 if None, will get by default rule(TODO: api key doc).
             page (int, optional): Page number. Defaults to 1.
             page_size (int, optional): Items per page. Defaults to 10.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: The fine-tune jobs in the result.
         """
-        return super().list(page, page_size, api_key, **kwargs)
+        return super().list(page,
+                            page_size,
+                            api_key,
+                            workspace=workspace,
+                            **kwargs)
 
     @classmethod
     def get(cls,
             file_id: str,
             api_key: str = None,
+            workspace: str = None,
             **kwargs) -> DashScopeAPIResponse:
         """Get the file info.
 
         Args:
             file_id (str): The file id.
             api_key (str, optional): The api key. Defaults to None.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: The job info
         """
-        return super().get(file_id, api_key, **kwargs)
+        return super().get(file_id, api_key, workspace=workspace, **kwargs)
 
     @classmethod
     def delete(cls,
                file_id: str,
                api_key: str = None,
+               workspace: str = None,
                **kwargs) -> DashScopeAPIResponse:
         """Delete uploaded file.
 
         Args:
             file_id (str): The file id want to delete.
             api_key (str, optional): The api key. Defaults to None.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: Delete result.
         """
-        return super().delete(file_id, api_key, **kwargs)
+        return super().delete(file_id, api_key, workspace=workspace, **kwargs)

dashscope/io/input_output.py

@@ -1,3 +1,5 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 import base64
 import io
 from typing import Generator

dashscope/model.py

@@ -1,3 +1,5 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 from dashscope.api_entities.dashscope_response import DashScopeAPIResponse
 from dashscope.client.base_api import GetMixin, ListMixin
 
@@ -9,23 +11,26 @@ class Model(ListMixin, GetMixin):
     def get(cls,
             name: str,
             api_key: str = None,
+            workspace: str = None,
             **kwargs) -> DashScopeAPIResponse:
         """Get the model information.
 
         Args:
             name (str): The model name.
             api_key (str, optional): The api key. Defaults to None.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: The model information.
         """
-        return super().get(name, api_key, **kwargs)
+        return super().get(name, api_key, workspace=workspace, **kwargs)
 
     @classmethod
     def list(cls,
              page=1,
              page_size=10,
              api_key: str = None,
+             workspace: str = None,
              **kwargs) -> DashScopeAPIResponse:
         """List models.
 
@@ -37,4 +42,8 @@ class Model(ListMixin, GetMixin):
         Returns:
             DashScopeAPIResponse: The models.
         """
-        return super().list(api_key, page, page_size, **kwargs)
+        return super().list(api_key,
+                            page,
+                            page_size,
+                            workspace=workspace,
+                            **kwargs)

dashscope/models.py

@@ -0,0 +1,43 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
+from dashscope.api_entities.dashscope_response import DashScopeAPIResponse
+from dashscope.client.base_api import GetMixin, ListMixin
+
+
+class Models(ListMixin, GetMixin):
+    SUB_PATH = 'models'
+
+    @classmethod
+    def get(cls,
+            name: str,
+            api_key: str = None,
+            **kwargs) -> DashScopeAPIResponse:
+        """Get the model information.
+
+        Args:
+            name (str): The model name.
+            api_key (str, optional): The api key. Defaults to None.
+            workspace (str): The dashscope workspace id.
+
+        Returns:
+            DashScopeAPIResponse: The model information.
+        """
+        return super().get(name, api_key, **kwargs)
+
+    @classmethod
+    def list(cls,
+             page=1,
+             page_size=10,
+             api_key: str = None,
+             **kwargs) -> DashScopeAPIResponse:
+        """List models.
+
+        Args:
+            api_key (str, optional): The api key
+            page (int, optional): Page number. Defaults to 1.
+            page_size (int, optional): Items per page. Defaults to 10.
+
+        Returns:
+            DashScopeAPIResponse: The models.
+        """
+        return super().list(page, page_size, api_key=api_key, **kwargs)

dashscope/multimodal/__init__.py

@@ -0,0 +1,20 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
+from .tingwu import tingwu
+from .tingwu.tingwu import TingWu
+from .tingwu.tingwu_realtime import TingWuRealtime, TingWuRealtimeCallback
+
+from .multimodal_dialog import MultiModalDialog, MultiModalCallback
+from .dialog_state import DialogState
+from .multimodal_constants import *
+from .multimodal_request_params import *
+
+__all__ = [
+    'tingwu',
+    'TingWu',
+    'TingWuRealtime',
+    'TingWuRealtimeCallback',
+    'MultiModalDialog',
+    'MultiModalCallback',
+    'DialogState'
+]

dashscope/multimodal/dialog_state.py

@@ -0,0 +1,56 @@
+# dialog_state.py
+
+from enum import Enum
+
+
+class DialogState(Enum):
+    """
+    对话状态枚举类，定义了对话机器人可能处于的不同状态。
+
+    Attributes:
+        IDLE (str): 表示机器人处于空闲状态。
+        LISTENING (str): 表示机器人正在监听用户输入。
+        THINKING (str): 表示机器人正在思考。
+        RESPONDING (str): 表示机器人正在生成或回复中。
+    """
+    IDLE = 'Idle'
+    LISTENING = 'Listening'
+    THINKING = 'Thinking'
+    RESPONDING = 'Responding'
+
+
+class StateMachine:
+    """
+    状态机类，用于管理机器人的状态转换。
+
+    Attributes:
+        current_state (DialogState): 当前状态。
+    """
+
+    def __init__(self):
+        # 初始化状态机时设置初始状态为IDLE
+        self.current_state = DialogState.IDLE
+
+    def change_state(self, new_state: str) -> None:
+        """
+        更改当前状态到指定的新状态。
+
+        Args:
+            new_state (str): 要切换到的新状态。
+
+        Raises:
+            ValueError: 如果尝试切换到一个无效的状态，则抛出此异常。
+        """
+        if new_state in [state.value for state in DialogState]:
+            self.current_state = DialogState(new_state)
+        else:
+            raise ValueError("无效的状态类型")
+
+    def get_current_state(self) -> DialogState:
+        """
+        获取当前状态。
+
+        Returns:
+            DialogState: 当前状态。
+        """
+        return self.current_state

dashscope/multimodal/multimodal_constants.py

@@ -0,0 +1,28 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+# -*- coding: utf-8 -*-
+
+# multimodal conversation request directive
+
+class RequestToRespondType:
+    TRANSCRIPT = 'transcript'
+    PROMPT = 'prompt'
+
+
+# multimodal conversation response directive
+RESPONSE_NAME_TASK_STARTED = "task-started"
+RESPONSE_NAME_RESULT_GENERATED = "result-generated"
+RESPONSE_NAME_TASK_FINISHED = "task-finished"
+
+RESPONSE_NAME_TASK_FAILED = "TaskFailed"
+RESPONSE_NAME_STARTED = "Started"
+RESPONSE_NAME_STOPPED = "Stopped"
+RESPONSE_NAME_STATE_CHANGED = "DialogStateChanged"
+RESPONSE_NAME_REQUEST_ACCEPTED = "RequestAccepted"
+RESPONSE_NAME_SPEECH_STARTED = "SpeechStarted"
+RESPONSE_NAME_SPEECH_ENDED = "SpeechEnded"  # 服务端检测到asr语音尾点时下发此事件,可选事件
+RESPONSE_NAME_RESPONDING_STARTED = "RespondingStarted"  # AI语音应答开始，sdk要准备接收服务端下发的语音数据
+RESPONSE_NAME_RESPONDING_ENDED = "RespondingEnded"  # AI语音应答结束
+RESPONSE_NAME_SPEECH_CONTENT = "SpeechContent"  # 用户语音识别出的文本，流式全量输出
+RESPONSE_NAME_RESPONDING_CONTENT = "RespondingContent"  # 统对外输出的文本，流式全量输出
+RESPONSE_NAME_ERROR = "Error"  # 服务端对话中报错
+RESPONSE_NAME_HEART_BEAT = "HeartBeat"  # 心跳消息