-
Notifications
You must be signed in to change notification settings - Fork 5.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat:add tts-streaming config and future #5492
Conversation
@charli117 Please review this pr as well. |
@crazywoola ic-xu is a colleague of our team, and the version submitted now is the version we are currently using. There is also an optimization point that requires adding a blocking mode output logic for models that are not compatible and do not support streaming tts. We have communicated with each other and merged later |
…e of streaming to implement TTS functionality
It has been extracted as a separate module, and processing logic has been added to the corresponding module in GenerateTask Pipeline. process(). At the same time, the expiration time has been set to be small. At the same time, Redis will not store all message objects, but only a portion of the required data, so the data will not be very large
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
See comments
I tested the manual play mode and found it to be unresponsive, with the chrome console reporting the error message as:
|
Maybe it's because they've all switched to streaming mode now, and the web front-end isn't compatible with manual playback mode. |
Could you provide a test browser? We are using Chrome for testing, which works well and has a fast response speed. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ops, now in the basic chatbot
, clicking the play button does not send a request.
Additionally, if the TTS function is enabled for the first time in the Chatflow
, clicking the play button will result in an error. This might be due to using a draft version instead of a published one. The error message is as follows:
Traceback (most recent call last):
File "/Users/takatost/Projects/dify/api/controllers/console/app/audio.py", line 97, in post
text_to_speech = app_model.workflow.features_dict.get('text_to_speech')
AttributeError: 'NoneType' object has no attribute 'features_dict'
…w, clicking the play button will result in an error.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
thx!
* refs/heads/main: (51 commits) feat: tailwind related improvement (#6085) feat: support AnalyticDB vector store (#5586) feat:add tts-streaming config and future (#5492) Feat: add index bar to select tool panel of workflow (#6066) bump to 0.6.13 (#6078) fix: Inconsistency Between Actual and Debug Input Variables (#6055) refactor: revamp picker block (#4227) chore: remove tsne unused code (#6077) fix: relative in overflow div (#5998) chore(action): move docker login above Set up QEMU in build-push action workflow (#6073) remove clunky welcome message (#6069) feat: add request_params field to jina_reader tool (#5610) fix azure stream download (#6063) chore: hide tracing introduce detail (#6049) Address the issue of the absence of poetry in the development container. (#6036) Fix authorization header validation to handle bearer types correctly - "authorization config header is required" error (#6040) Fix/6034 get random order of categories in explore and workflow is missing in zh hant (#6043) Modify slack webhook url validation to allow workflow (#6041) (#6042) fix(configs): Update pydantic settings in config files (#6023) Fix/incorrect parameter extractor memory (#6038) ...
Description
Add TTS streaming configuration and feature support, add text to speech in the feature configuration options, click on configure to start
Fixes #5251
Type of Change
How Has This Been Tested?
Simply configure the streaming TTS option and test it