feat: 增加阿里百炼语音模型

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/aliyun_bai_lian_model_provider.py

@@ -13,13 +13,26 @@ from setting.models_provider.base_model_provider import ModelProvideInfo, ModelT
     ModelInfoManage
 from setting.models_provider.impl.aliyun_bai_lian_model_provider.credential.reranker import \
     AliyunBaiLianRerankerCredential
+from setting.models_provider.impl.aliyun_bai_lian_model_provider.credential.stt import AliyunBaiLianSTTModelCredential
+from setting.models_provider.impl.aliyun_bai_lian_model_provider.credential.tts import AliyunBaiLianTTSModelCredential
 from setting.models_provider.impl.aliyun_bai_lian_model_provider.model.reranker import AliyunBaiLianReranker
+from setting.models_provider.impl.aliyun_bai_lian_model_provider.model.stt import AliyunBaiLianSpeechToText
+from setting.models_provider.impl.aliyun_bai_lian_model_provider.model.tts import AliyunBaiLianTextToSpeech
 from smartdoc.conf import PROJECT_DIR
 
 aliyun_bai_lian_model_credential = AliyunBaiLianRerankerCredential()
+aliyun_bai_lian_tts_model_credential = AliyunBaiLianTTSModelCredential()
+aliyun_bai_lian_stt_model_credential = AliyunBaiLianSTTModelCredential()
+
 model_info_list = [ModelInfo('gte-rerank',
                              '阿里巴巴通义实验室开发的GTE-Rerank文本排序系列模型，开发者可以通过LlamaIndex框架进行集成高质量文本检索、排序。',
-                             ModelTypeConst.RERANKER, aliyun_bai_lian_model_credential, AliyunBaiLianReranker)
+                             ModelTypeConst.RERANKER, aliyun_bai_lian_model_credential, AliyunBaiLianReranker),
+                   ModelInfo('paraformer-realtime-v2',
+                             '中文（含粤语等各种方言）、英文、日语、韩语支持多个语种自由切换',
+                             ModelTypeConst.STT, aliyun_bai_lian_stt_model_credential, AliyunBaiLianSpeechToText),
+                   ModelInfo('cosyvoice-v1',
+                             'CosyVoice基于新一代生成式语音大模型，能根据上下文预测情绪、语调、韵律等，具有更好的拟人效果',
+                             ModelTypeConst.TTS, aliyun_bai_lian_tts_model_credential, AliyunBaiLianTextToSpeech),
                    ]
 
 model_info_manage = ModelInfoManage.builder().append_model_info_list(model_info_list).build()

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/credential/stt.py

@@ -0,0 +1,42 @@
+# coding=utf-8
+
+from typing import Dict
+
+from common import forms
+from common.exception.app_exception import AppApiException
+from common.forms import BaseForm
+from setting.models_provider.base_model_provider import BaseModelCredential, ValidCode
+
+
+class AliyunBaiLianSTTModelCredential(BaseForm, BaseModelCredential):
+    api_key = forms.PasswordInputField("API Key", required=True)
+
+    def is_valid(self, model_type: str, model_name, model_credential: Dict[str, object], provider,
+                 raise_exception=False):
+        model_type_list = provider.get_model_type_list()
+        if not any(list(filter(lambda mt: mt.get('value') == model_type, model_type_list))):
+            raise AppApiException(ValidCode.valid_error.value, f'{model_type} 模型类型不支持')
+
+        for key in ['api_key']:
+            if key not in model_credential:
+                if raise_exception:
+                    raise AppApiException(ValidCode.valid_error.value, f'{key} 字段为必填字段')
+                else:
+                    return False
+        try:
+            model = provider.get_model(model_type, model_name, model_credential)
+            model.check_auth()
+        except Exception as e:
+            if isinstance(e, AppApiException):
+                raise e
+            if raise_exception:
+                raise AppApiException(ValidCode.valid_error.value, f'校验失败,请检查参数是否正确: {str(e)}')
+            else:
+                return False
+        return True
+
+    def encryption_dict(self, model: Dict[str, object]):
+        return {**model, 'api_key': super().encryption(model.get('api_key', ''))}
+
+    def get_model_params_setting_form(self, model_name):
+        pass

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/credential/tts.py

@@ -0,0 +1,43 @@
+# coding=utf-8
+
+from typing import Dict
+
+from common import forms
+from common.exception.app_exception import AppApiException
+from common.forms import BaseForm
+from setting.models_provider.base_model_provider import BaseModelCredential, ValidCode
+
+
+class AliyunBaiLianTTSModelCredential(BaseForm, BaseModelCredential):
+    api_key = forms.PasswordInputField("API Key", required=True)
+
+    def is_valid(self, model_type: str, model_name, model_credential: Dict[str, object], provider,
+                 raise_exception=False):
+        model_type_list = provider.get_model_type_list()
+        if not any(list(filter(lambda mt: mt.get('value') == model_type, model_type_list))):
+            raise AppApiException(ValidCode.valid_error.value, f'{model_type} 模型类型不支持')
+
+        for key in ['api_key']:
+            if key not in model_credential:
+                if raise_exception:
+                    raise AppApiException(ValidCode.valid_error.value, f'{key} 字段为必填字段')
+                else:
+                    return False
+        try:
+            model = provider.get_model(model_type, model_name, model_credential)
+            model.check_auth()
+        except Exception as e:
+            if isinstance(e, AppApiException):
+                raise e
+            if raise_exception:
+                raise AppApiException(ValidCode.valid_error.value, f'校验失败,请检查参数是否正确: {str(e)}')
+            else:
+                return False
+        return True
+
+    def encryption_dict(self, model: Dict[str, object]):
+        return {**model, 'api_key': super().encryption(model.get('api_key', ''))}
+
+
+    def get_model_params_setting_form(self, model_name):
+        pass

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/model/stt.py

@@ -0,0 +1,63 @@
+import os
+import tempfile
+from typing import Dict
+
+import dashscope
+from dashscope.audio.asr import (Recognition)
+
+from setting.models_provider.base_model_provider import MaxKBBaseModel
+from setting.models_provider.impl.base_stt import BaseSpeechToText
+
+
+class AliyunBaiLianSpeechToText(MaxKBBaseModel, BaseSpeechToText):
+    api_key: str
+    model: str
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.api_key = kwargs.get('api_key')
+        self.model = kwargs.get('model')
+
+    @staticmethod
+    def new_instance(model_type, model_name, model_credential: Dict[str, object], **model_kwargs):
+        optional_params = {}
+        if 'max_tokens' in model_kwargs and model_kwargs['max_tokens'] is not None:
+            optional_params['max_tokens'] = model_kwargs['max_tokens']
+        if 'temperature' in model_kwargs and model_kwargs['temperature'] is not None:
+            optional_params['temperature'] = model_kwargs['temperature']
+        return AliyunBaiLianSpeechToText(
+            model=model_name,
+            api_key=model_credential.get('api_key'),
+            **optional_params,
+        )
+
+    def check_auth(self):
+        cwd = os.path.dirname(os.path.abspath(__file__))
+        with open(f'{cwd}/iat_mp3_16k.mp3', 'rb') as f:
+            self.speech_to_text(f)
+
+    def speech_to_text(self, audio_file):
+        dashscope.api_key = self.api_key
+        recognition = Recognition(model=self.model,
+                                  format='mp3',
+                                  sample_rate=16000,
+                                  callback=None)
+        with tempfile.NamedTemporaryFile(delete=False) as temp_file:
+            # 将上传的文件保存到临时文件中
+            temp_file.write(audio_file.read())
+            # 获取临时文件的路径
+            temp_file_path = temp_file.name
+
+        try:
+            # 识别临时文件
+            result = recognition.call(temp_file_path)
+            text = ''
+            if result.status_code == 200:
+                for sentence in result.get_sentence():
+                    text += sentence['text']
+                return text
+            else:
+                raise Exception('Error: ', result.message)
+        finally:
+            # 删除临时文件
+            os.remove(temp_file_path)

apps/setting/models_provider/impl/aliyun_bai_lian_model_provider/model/tts.py

@@ -0,0 +1,43 @@
+from typing import Dict
+
+import dashscope
+from dashscope.audio.tts_v2 import *
+
+from setting.models_provider.base_model_provider import MaxKBBaseModel
+from setting.models_provider.impl.base_tts import BaseTextToSpeech
+
+
+class AliyunBaiLianTextToSpeech(MaxKBBaseModel, BaseTextToSpeech):
+    api_key: str
+    model: str
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.api_key = kwargs.get('api_key')
+        self.model = kwargs.get('model')
+
+    @staticmethod
+    def new_instance(model_type, model_name, model_credential: Dict[str, object], **model_kwargs):
+        optional_params = {}
+        if 'max_tokens' in model_kwargs and model_kwargs['max_tokens'] is not None:
+            optional_params['max_tokens'] = model_kwargs['max_tokens']
+        if 'temperature' in model_kwargs and model_kwargs['temperature'] is not None:
+            optional_params['temperature'] = model_kwargs['temperature']
+        return AliyunBaiLianTextToSpeech(
+            model=model_name,
+            api_key=model_credential.get('api_key'),
+            **optional_params,
+        )
+
+    def check_auth(self):
+        self.text_to_speech('你好')
+
+    def text_to_speech(self, text):
+        voice = "longxiaochun"
+        dashscope.api_key = self.api_key
+        synthesizer = SpeechSynthesizer(model=self.model, voice=voice)
+        audio = synthesizer.call(text)
+        if type(audio) == str:
+            print(audio)
+            raise Exception(audio)
+        return audio