Hugging Face's logo

Spaces:
sohojoe
/
project_charles
Sleeping

App Files Community
project_charles

Commit History

improve the app layout a little
e2846c4

sohojoe commited on

Update requirements.txt
f2296f2

sohojoe commited on

add ray
1e20df9

sohojoe commited on

fix type issue
9b7ec10

sohojoe commited on

change streamlit version
73d75a1

sohojoe commited on

add error handling to openai api
156e5b3

sohojoe commited on

default to v1 voice
b49fa2f

sohojoe commited on

add missing async / await
0d27fd9

sohojoe commited on

move pipeline to legacy
572f2e6

sohojoe commited on

fix: potential bug in SpeechToSpeakerActor
610cc37

sohojoe commited on

refactor - use app.py
da9d44b

sohojoe commited on

refactor
1a63d97

sohojoe commited on

reduce complexity by removing AudioStreamProcessor from StreamingChatService
e9d6e62

sohojoe commited on

fix async issues and flush ffmpeg actor
df0ea75

sohojoe commited on

add flush_output_queue
ae52b65

sohojoe commited on

prep for eleven_english_v2
6f108ca

sohojoe commited on

ensure echo cancelation
de63932

sohojoe commited on

attempt (request) better audio. not sure it makes much of a difference
8332d4f

sohojoe commited on

first version streaming via webrtc but quality sucks
d91a673

sohojoe commited on

refactor so we have in and then i can add out
ac35a95

sohojoe commited on

refactor ffmeg_concerter_actor.py to its own class
c4fe843

sohojoe commited on

updated version that uses ray queues so it doesn't block.
681abd4

sohojoe commited on

basic version that is working
6dfcd3a

sohojoe commited on

debug_app.py which choppy audio
d0639dc

sohojoe commited on

pickle the frames so easier to debug downstream
32ed25f

sohojoe commited on

4096 chunk sizes seams to be more robust
24d9393

sohojoe commited on

add mp3 format hint
7efb19e

sohojoe commited on

some debug code
f615884

sohojoe commited on

fix: should be raise, now throw
dc249ac

sohojoe commited on

should not leave on debug text
187e31c

sohojoe commited on

add cancel and fix a bunch of async problems
afc1e50

sohojoe commited on

create respond to prompt actor
bcea2ea

sohojoe commited on

migrated speech to text to an actor
98ec0ec

sohojoe commited on

fix use of ray.get inside aync
768e92a

sohojoe commited on

Charles running between two applications
ad67495

sohojoe commited on

start of moving to ray
62a21bd

sohojoe commited on

prep for charicter
ff684ee

sohojoe commited on

prep for video frames
ad9371a

sohojoe commited on

basic POC
162d5c8

sohojoe commited on

refactor app.py to run as async
9740bc5

sohojoe commited on

created chat_pipeline
730fe87

sohojoe commited on

add: pipelines nodes can now spawn one to many jobs via yield
8c9e2db

sohojoe commited on

create a pipeline class with basic test coverage
ed232fa

sohojoe commited on

create a node class to help things be a bit more robust
de8cac1

sohojoe commited on

added pipeline test for a simple pipeline
80eea9e

sohojoe commited on

added debug time to show the time multiple of actual audio time to streaming time
31bfb8d

sohojoe commited on

add some debug code that suggest that we are still been buffered somehow
ac171cf

sohojoe commited on

add _safe_enqueue_text_to_speak
1fe79e2

sohojoe commited on

fix issues with _should_we_send_to_voice and add a test
12f4772

sohojoe commited on

fix: close would lock
245e7f9

sohojoe commited on