Files
dify-plugin-sdks/python/dify_plugin/interfaces/model/speech2text_model.py
immortal-wm 2fc9753673 fix: corrected latency accuracy (#189)
* fix latency calc

* feat: add timing context management to AIModel for latency tracking

- Introduced a new TimingContextRaceConditionError to handle race conditions in multi-threaded environments.
- Implemented a timing_context method in AIModel to track request timing and prevent concurrent usage.
- Updated various model classes (LargeLanguageModel, ModerationModel, RerankModel, Speech2TextModel, TextEmbeddingModel, TTSModel) to utilize the new timing context for latency calculations.
- Added tests to validate timing context behavior under concurrent and sequential usage scenarios.

* feat: implement ModelFactory for dynamic model instance creation

- Added ModelFactory class to generate stateless model instances based on provider configurations and model classes.
- Updated PluginRegistration to utilize ModelFactory for managing model instances, enhancing code organization and maintainability.

* feat: add unit tests for model registry and mock model provider

- Introduced a new test file for validating model registry functionality.
- Implemented mock classes for model provider and LLM to facilitate testing.
- Added tests to ensure correct model instance retrieval from the registry.

* apply ruff

---------

Co-authored-by: Yeuoly <admin@srmxy.cn>
2025-09-16 16:15:20 +08:00

73 lines
2.3 KiB
Python

import os
from abc import abstractmethod
from typing import IO
from pydantic import ConfigDict
from dify_plugin.entities.model import ModelType
from dify_plugin.interfaces.model.ai_model import AIModel
class Speech2TextModel(AIModel):
"""
Model class for speech2text model.
"""
model_type: ModelType = ModelType.SPEECH2TEXT
# pydantic configs
model_config = ConfigDict(protected_namespaces=())
############################################################
# Methods that can be implemented by plugin #
############################################################
@abstractmethod
def _invoke(self, model: str, credentials: dict, file: IO[bytes], user: str | None = None) -> str:
"""
Invoke large language model
:param model: model name
:param credentials: model credentials
:param file: audio file
:param user: unique user id
:return: text for given audio file
"""
raise NotImplementedError
############################################################
# For plugin implementation use only #
############################################################
def _get_demo_file_path(self) -> str:
"""
Get demo file for given model
:return: demo file
"""
# Get the directory of the current file
current_dir = os.path.dirname(os.path.abspath(__file__))
# Construct the path to the audio file
return os.path.join(current_dir, "audio.mp3")
############################################################
# For executor use only #
############################################################
def invoke(self, model: str, credentials: dict, file: IO[bytes], user: str | None = None) -> str:
"""
Invoke large language model
:param model: model name
:param credentials: model credentials
:param file: audio file
:param user: unique user id
:return: text for given audio file
"""
with self.timing_context():
try:
return self._invoke(model, credentials, file, user)
except Exception as e:
raise self._transform_invoke_error(e) from e