selfhostyourtech/reflector
2
0
Fork 0
mirror of https://github.com/Monadical-SAS/reflector.git synced 2025-12-21 04:39:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

code cleanup

Browse Source
This commit is contained in:
Gokul Mohanarangan
2023-07-11 12:09:30 +05:30
parent b7fbfb2a54
commit 8e9cd6c568
15 changed files with 249 additions and 215 deletions
Download Patch File Download Diff File Expand all files Collapse all files

125
README.md
View File

@@ -1,32 +1,34 @@
# Reflector # Reflector
This is the code base for the Reflector demo (formerly called agenda-talk-diff) for the leads : Troy Web Consulting panel (A Chat with AWS about AI: Real AI/ML AWS projects and what you should know) on 6/14 at 430PM. This is the code base for the Reflector demo (formerly called agenda-talk-diff) for the leads : Troy Web Consulting
panel (A Chat with AWS about AI: Real AI/ML AWS projects and what you should know) on 6/14 at 430PM.
The target deliverable is a local-first live transcription and visualization tool to compare a discussion's target agenda/objectives to the actual discussion live.
The target deliverable is a local-first live transcription and visualization tool to compare a discussion's target
agenda/objectives to the actual discussion live.
**S3 bucket:** **S3 bucket:**
Everything you need for S3 is already configured in config.ini. Only edit it if you need to change it deliberately. Everything you need for S3 is already configured in config.ini. Only edit it if you need to change it deliberately.
S3 bucket name is mentioned in config.ini. All transfers will happen between this bucket and the local computer where the S3 bucket name is mentioned in config.ini. All transfers will happen between this bucket and the local computer where
script is run. You need AWS_ACCESS_KEY / AWS_SECRET_KEY to authenticate your calls to S3 (done in config.ini). the
script is run. You need AWS_ACCESS_KEY / AWS_SECRET_KEY to authenticate your calls to S3 (done in config.ini).
For AWS S3 Web UI, For AWS S3 Web UI,
1) Login to AWS management console. 1) Login to AWS management console.
2) Search for S3 in the search bar at the top. 2) Search for S3 in the search bar at the top.
3) Navigate to list the buckets under the current account, if needed and choose your bucket [```reflector-bucket```] 3) Navigate to list the buckets under the current account, if needed and choose your bucket [```reflector-bucket```]
4) You should be able to see items in the bucket. You can upload/download files here directly. 4) You should be able to see items in the bucket. You can upload/download files here directly.
For CLI,
For CLI,
Refer to the FILE UTIL section below. Refer to the FILE UTIL section below.
**FILE UTIL MODULE:** **FILE UTIL MODULE:**
A file_util module has been created to upload/download files with AWS S3 bucket pre-configured using config.ini. A file_util module has been created to upload/download files with AWS S3 bucket pre-configured using config.ini.
Though not needed for the workflow, if you need to upload / download file, separately on your own, apart from the pipeline workflow in the script, you can do so by : Though not needed for the workflow, if you need to upload / download file, separately on your own, apart from the
pipeline workflow in the script, you can do so by :
Upload: Upload:
@@ -39,37 +41,37 @@ Download:
If you want to access the S3 artefacts, from another machine, you can either use the python file_util with the commands If you want to access the S3 artefacts, from another machine, you can either use the python file_util with the commands
mentioned above or simply use the GUI of AWS Management Console. mentioned above or simply use the GUI of AWS Management Console.
To setup,
To setup,
1) Check values in config.ini file. Specifically add your OPENAI_APIKEY if you plan to use OpenAI API requests. 1) Check values in config.ini file. Specifically add your OPENAI_APIKEY if you plan to use OpenAI API requests.
2) Run ``` export KMP_DUPLICATE_LIB_OK=True``` in Terminal. [This is taken care of in code, but not reflecting, Will fix this issue later.] 2) Run ``` export KMP_DUPLICATE_LIB_OK=True``` in
Terminal. [This is taken care of in code, but not reflecting, Will fix this issue later.]
NOTE: If you don't have portaudio installed already, run ```brew install portaudio``` NOTE: If you don't have portaudio installed already, run ```brew install portaudio```
3) Run the script setup_depedencies.sh. 3) Run the script setup_depedencies.sh.
``` chmod +x setup_dependencies.sh ``` ``` chmod +x setup_dependencies.sh ```
``` sh setup_dependencies.sh <ENV>``` ``` sh setup_dependencies.sh <ENV>```
ENV refers to the intended environment for JAX. JAX is available in several
ENV refers to the intended environment for JAX. JAX is available in several variants, [CPU | GPU | Colab TPU | Google Cloud TPU] variants, [CPU | GPU | Colab TPU | Google Cloud TPU]
```ENV``` is :
cpu -> JAX CPU installation
cuda11 -> JAX CUDA 11.x version ```ENV``` is :
cuda12 -> JAX CUDA 12.x version (Core Weave has CUDA 12 version, can check with ```nvidia-smi```) cpu -> JAX CPU installation
cuda11 -> JAX CUDA 11.x version
cuda12 -> JAX CUDA 12.x version (Core Weave has CUDA 12 version, can check with ```nvidia-smi```)
```sh setup_dependencies.sh cuda12``` ```sh setup_dependencies.sh cuda12```
4) If not already done, install ffmpeg. ```brew install ffmpeg``` 4) If not already done, install ffmpeg. ```brew install ffmpeg```
For NLTK SSL error, check [here](https://stackoverflow.com/questions/38916452/nltk-download-ssl-certificate-verify-failed) For NLTK SSL error,
check [here](https://stackoverflow.com/questions/38916452/nltk-download-ssl-certificate-verify-failed)
5) Run the Whisper-JAX pipeline. Currently, the repo can take a Youtube video and transcribes/summarizes it. 5) Run the Whisper-JAX pipeline. Currently, the repo can take a Youtube video and transcribes/summarizes it.
@@ -79,83 +81,92 @@ You can even run it on local file or a file in your configured S3 bucket.
``` python3 whisjax.py "startup.mp4"``` ``` python3 whisjax.py "startup.mp4"```
The script will take care of a few cases like youtube file, local file, video file, audio-only file, The script will take care of a few cases like youtube file, local file, video file, audio-only file,
file in S3, etc. If local file is not present, it can automatically take the file from S3. file in S3, etc. If local file is not present, it can automatically take the file from S3.
**OFFLINE WORKFLOW:** **OFFLINE WORKFLOW:**
1) Specify the input source file] from a local, youtube link or upload to S3 if needed and pass it as input to the script.If the source file is in 1) Specify the input source file] from a local, youtube link or upload to S3 if needed and pass it as input to the
script.If the source file is in
```.m4a``` format, it will get converted to ```.mp4``` automatically. ```.m4a``` format, it will get converted to ```.mp4``` automatically.
2) Keep the agenda header topics in a local file named ```agenda-headers.txt```. This needs to be present where the script is run. 2) Keep the agenda header topics in a local file named ```agenda-headers.txt```. This needs to be present where the
script is run.
This version of the pipeline compares covered agenda topics using agenda headers in the following format. This version of the pipeline compares covered agenda topics using agenda headers in the following format.
1) ```agenda_topic : <short description>``` 1) ```agenda_topic : <short description>```
3) Check all the values in ```config.ini```. You need to predefine 2 categories for which you need to scatter plot the 3) Check all the values in ```config.ini```. You need to predefine 2 categories for which you need to scatter plot the
topic modelling visualization in the config file. This is the default visualization. But, from the dataframe artefact called topic modelling visualization in the config file. This is the default visualization. But, from the dataframe artefact
```df_<timestamp>.pkl``` , you can load the df and choose different topics to plot. You can filter using certain words to search for the called
```df_<timestamp>.pkl``` , you can load the df and choose different topics to plot. You can filter using certain
words to search for the
transcriptions and you can see the top influencers and characteristic in each topic we have chosen to plot in the transcriptions and you can see the top influencers and characteristic in each topic we have chosen to plot in the
interactive HTML document. I have added a new jupyter notebook that gives the base template to play around with, named interactive HTML document. I have added a new jupyter notebook that gives the base template to play around with,
named
```Viz_experiments.ipynb```. ```Viz_experiments.ipynb```.
4) Run the script. The script automatically transcribes, summarizes and creates a scatter plot of words & topics in the form of an interactive 4) Run the script. The script automatically transcribes, summarizes and creates a scatter plot of words & topics in the
HTML file, a sample word cloud and uploads them to the S3 bucket form of an interactive
HTML file, a sample word cloud and uploads them to the S3 bucket
5) Additional artefacts pushed to S3: 5) Additional artefacts pushed to S3:
1) HTML visualization file 1) HTML visualization file
2) pandas df in pickle format for others to collaborate and make their own visualizations 2) pandas df in pickle format for others to collaborate and make their own visualizations
3) Summary, transcript and transcript with timestamps file in text format. 3) Summary, transcript and transcript with timestamps file in text format.
The script also creates 2 types of mappings. The script also creates 2 types of mappings.
1) Timestamp -> The top 2 matched agenda topic 1) Timestamp -> The top 2 matched agenda topic
2) Topic -> All matched timestamps in the transcription 2) Topic -> All matched timestamps in the transcription
Other visualizations can be planned based on available artefacts or new ones can be created. Refer the section ```Viz-experiments```.
Other visualizations can be planned based on available artefacts or new ones can be created. Refer the
section ```Viz-experiments```.
**Visualization experiments:** **Visualization experiments:**
This is a jupyter notebook playground with template instructions on handling the metadata and data artefacts generated from the This is a jupyter notebook playground with template instructions on handling the metadata and data artefacts generated
pipeline. Follow the instructions given and tweak your own logic into it or use it as a playground to experiment libraries and from the
pipeline. Follow the instructions given and tweak your own logic into it or use it as a playground to experiment
libraries and
visualizations on top of the metadata. visualizations on top of the metadata.
**WHISPER-JAX REALTIME TRANSCRIPTION PIPELINE:** **WHISPER-JAX REALTIME TRANSCRIPTION PIPELINE:**
We also support a provision to perform real-time transcripton using whisper-jax pipeline. But, there are We also support a provision to perform real-time transcripton using whisper-jax pipeline. But, there are
a few pre-requisites before you run it on your local machine. The instructions are for a few pre-requisites before you run it on your local machine. The instructions are for
configuring on a MacOS. configuring on a MacOS.
We need to way to route audio from an application opened via the browser, ex. "Whereby" and audio from your local We need to way to route audio from an application opened via the browser, ex. "Whereby" and audio from your local
microphone input which you will be using for speaking. We use [Blackhole](https://github.com/ExistentialAudio/BlackHole). microphone input which you will be using for speaking. We
use [Blackhole](https://github.com/ExistentialAudio/BlackHole).
1) Install Blackhole-2ch (2 ch is enough) by 1 of 2 options listed. 1) Install Blackhole-2ch (2 ch is enough) by 1 of 2 options listed.
2) Setup [Aggregate device](https://github.com/ExistentialAudio/BlackHole/wiki/Aggregate-Device) to route web audio and 2) Setup [Aggregate device](https://github.com/ExistentialAudio/BlackHole/wiki/Aggregate-Device) to route web audio and
local microphone input. local microphone input.
Be sure to mirror the settings given ![here](./images/aggregate_input.png) Be sure to mirror the settings given ![here](./images/aggregate_input.png)
3) Setup [Multi-Output device](https://github.com/ExistentialAudio/BlackHole/wiki/Multi-Output-Device) 3) Setup [Multi-Output device](https://github.com/ExistentialAudio/BlackHole/wiki/Multi-Output-Device)
Refer ![here](./images/multi-output.png) Refer ![here](./images/multi-output.png)
4) Set the aggregator input device name created in step 2 in config.ini as ```BLACKHOLE_INPUT_AGGREGATOR_DEVICE_NAME``` 4) Set the aggregator input device name created in step 2 in config.ini as ```BLACKHOLE_INPUT_AGGREGATOR_DEVICE_NAME```
5) Then goto ``` System Preferences -> Sound ``` and choose the devices created from the Output and 5) Then goto ``` System Preferences -> Sound ``` and choose the devices created from the Output and
Input tabs. Input tabs.
6) The input from your local microphone, the browser run meeting should be aggregated into one virtual stream to listen to 6) The input from your local microphone, the browser run meeting should be aggregated into one virtual stream to listen
and the output should be fed back to your specified output devices if everything is configured properly. Check this to
before trying out the trial. and the output should be fed back to your specified output devices if everything is configured properly. Check this
before trying out the trial.
**Permissions:** **Permissions:**
You may have to add permission for "Terminal"/Code Editors [Pycharm/VSCode, etc.] microphone access to record audio in You may have to add permission for "Terminal"/Code Editors [Pycharm/VSCode, etc.] microphone access to record audio in
```System Preferences -> Privacy & Security -> Microphone```, ```System Preferences -> Privacy & Security -> Microphone```,
```System Preferences -> Privacy & Security -> Accessibility```, ```System Preferences -> Privacy & Security -> Accessibility```,
```System Preferences -> Privacy & Security -> Input Monitoring```. ```System Preferences -> Privacy & Security -> Input Monitoring```.
From the reflector root folder, From the reflector root folder,
run ```python3 whisjax_realtime.py``` run ```python3 whisjax_realtime.py```
The transcription text should be written to ```real_time_transcription_<timestamp>.txt```. The transcription text should be written to ```real_time_transcription_<timestamp>.txt```.
NEXT STEPS: NEXT STEPS:
1) Create a RunPod setup for this feature (mentioned in 1 & 2) and test it end-to-end 1) Create a RunPod setup for this feature (mentioned in 1 & 2) and test it end-to-end

30
client.py
View File

@@ -1,33 +1,33 @@
import argparse import argparse
import asyncio import asyncio
import signal import signal
from utils.log_utils import logger
from aiortc.contrib.signaling import (add_signaling_arguments, from aiortc.contrib.signaling import (add_signaling_arguments,
create_signaling) create_signaling)
from stream_client import StreamClient from stream_client import StreamClient
from utils.log_utils import logger
async def main(): async def main():
parser = argparse.ArgumentParser(description="Data channels ping/pong") parser = argparse.ArgumentParser(description="Data channels ping/pong")
parser.add_argument( parser.add_argument(
"--url", type=str, nargs="?", default="http://127.0.0.1:1250/offer" "--url", type=str, nargs="?", default="http://127.0.0.1:1250/offer"
) )
parser.add_argument( parser.add_argument(
"--ping-pong", "--ping-pong",
help="Benchmark data channel with ping pong", help="Benchmark data channel with ping pong",
type=eval, type=eval,
choices=[True, False], choices=[True, False],
default="False", default="False",
) )
parser.add_argument( parser.add_argument(
"--play-from", "--play-from",
type=str, type=str,
default="", default="",
) )
add_signaling_arguments(parser) add_signaling_arguments(parser)
@@ -54,14 +54,14 @@ async def main():
loop = asyncio.get_event_loop() loop = asyncio.get_event_loop()
for s in signals: for s in signals:
loop.add_signal_handler( loop.add_signal_handler(
s, lambda s=s: asyncio.create_task(shutdown(s, loop))) s, lambda s=s: asyncio.create_task(shutdown(s, loop)))
# Init client # Init client
sc = StreamClient( sc = StreamClient(
signaling=signaling, signaling=signaling,
url=args.url, url=args.url,
play_from=args.play_from, play_from=args.play_from,
ping_pong=args.ping_pong ping_pong=args.ping_pong
) )
await sc.start() await sc.start()
print("Stream client started") print("Stream client started")

30
config.ini
View File

@@ -1,22 +1,22 @@
[DEFAULT] [DEFAULT]
# Set exception rule for OpenMP error to allow duplicate lib initialization # Set exception rule for OpenMP error to allow duplicate lib initialization
KMP_DUPLICATE_LIB_OK=TRUE KMP_DUPLICATE_LIB_OK = TRUE
# Export OpenAI API Key # Export OpenAI API Key
OPENAI_APIKEY= OPENAI_APIKEY =
# Export Whisper Model Size # Export Whisper Model Size
WHISPER_MODEL_SIZE=tiny WHISPER_MODEL_SIZE = tiny
WHISPER_REAL_TIME_MODEL_SIZE=tiny WHISPER_REAL_TIME_MODEL_SIZE = tiny
# AWS config # AWS config
AWS_ACCESS_KEY=***REMOVED*** AWS_ACCESS_KEY = ***REMOVED***
AWS_SECRET_KEY=***REMOVED*** AWS_SECRET_KEY = ***REMOVED***
BUCKET_NAME='reflector-bucket' BUCKET_NAME = 'reflector-bucket'
# Summarizer config # Summarizer config
SUMMARY_MODEL=facebook/bart-large-cnn SUMMARY_MODEL = facebook/bart-large-cnn
INPUT_ENCODING_MAX_LENGTH=1024 INPUT_ENCODING_MAX_LENGTH = 1024
MAX_LENGTH=2048 MAX_LENGTH = 2048
BEAM_SIZE=6 BEAM_SIZE = 6
MAX_CHUNK_LENGTH=1024 MAX_CHUNK_LENGTH = 1024
SUMMARIZE_USING_CHUNKS=YES SUMMARIZE_USING_CHUNKS = YES
# Audio device # Audio device
BLACKHOLE_INPUT_AGGREGATOR_DEVICE_NAME=aggregator BLACKHOLE_INPUT_AGGREGATOR_DEVICE_NAME = aggregator
AV_FOUNDATION_DEVICE_ID=2 AV_FOUNDATION_DEVICE_ID = 2

4
requirements.txt
View File

@@ -26,7 +26,7 @@ networkx==3.1
numba==0.57.0 numba==0.57.0
numpy==1.24.3 numpy==1.24.3
openai==0.27.7 openai==0.27.7
openai-whisper @ git+https://github.com/openai/whisper.git@248b6cb124225dd263bb9bd32d060b6517e067f8 openai-whisper@ git+https://github.com/openai/whisper.git@248b6cb124225dd263bb9bd32d060b6517e067f8
Pillow==9.5.0 Pillow==9.5.0
proglog==0.1.10 proglog==0.1.10
pytube==15.0.0 pytube==15.0.0
@@ -56,5 +56,5 @@ cached_property==1.5.2
stamina==23.1.0 stamina==23.1.0
httpx==0.24.1 httpx==0.24.1
sortedcontainers==2.4.0 sortedcontainers==2.4.0
openai-whisper @ git+https://github.com/openai/whisper.git@248b6cb124225dd263bb9bd32d060b6517e067f8 openai-whisper@ git+https://github.com/openai/whisper.git@248b6cb124225dd263bb9bd32d060b6517e067f8
https://github.com/yt-dlp/yt-dlp/archive/master.tar.gz https://github.com/yt-dlp/yt-dlp/archive/master.tar.gz

26
server_executor_cleaned.py
View File

@@ -15,7 +15,7 @@ from av import AudioFifo
from loguru import logger from loguru import logger
from whisper_jax import FlaxWhisperPipline from whisper_jax import FlaxWhisperPipline
from utils.server_utils import run_in_executor from utils.run_utils import run_in_executor
transcription = "" transcription = ""
@@ -44,10 +44,10 @@ def channel_send(channel, message):
if channel: if channel:
channel.send(message) channel.send(message)
print( print(
"Bytes handled :", "Bytes handled :",
total_bytes_handled, total_bytes_handled,
" Time : ", " Time : ",
datetime.datetime.now() - start_time, datetime.datetime.now() - start_time,
) )
@@ -86,12 +86,12 @@ class AudioStreamTrack(MediaStreamTrack):
audio_buffer.write(frame) audio_buffer.write(frame)
if local_frames := audio_buffer.read_many(256 * 960, partial=False): if local_frames := audio_buffer.read_many(256 * 960, partial=False):
whisper_result = run_in_executor( whisper_result = run_in_executor(
get_transcription, local_frames, executor=executor get_transcription, local_frames, executor=executor
) )
whisper_result.add_done_callback( whisper_result.add_done_callback(
lambda f: channel_send(data_channel, str(whisper_result.result())) lambda f: channel_send(data_channel, str(whisper_result.result()))
if (f.result()) if (f.result())
else None else None
) )
return frame return frame
@@ -140,10 +140,10 @@ async def offer(request):
answer = await pc.createAnswer() answer = await pc.createAnswer()
await pc.setLocalDescription(answer) await pc.setLocalDescription(answer)
return web.Response( return web.Response(
content_type="application/json", content_type="application/json",
text=json.dumps( text=json.dumps(
{"sdp": pc.localDescription.sdp, "type": pc.localDescription.type} { "sdp": pc.localDescription.sdp, "type": pc.localDescription.type }
), ),
) )

27
server_multithreaded.py
View File

@@ -1,5 +1,5 @@
import asyncio import asyncio
import configparser from utils.run_utils import config
import datetime import datetime
import io import io
import json import json
@@ -8,9 +8,9 @@ import threading
import uuid import uuid
import wave import wave
from concurrent.futures import ThreadPoolExecutor from concurrent.futures import ThreadPoolExecutor
from aiohttp import web
import jax.numpy as jnp import jax.numpy as jnp
from aiohttp import web
from aiortc import MediaStreamTrack, RTCPeerConnection, RTCSessionDescription from aiortc import MediaStreamTrack, RTCPeerConnection, RTCSessionDescription
from aiortc.contrib.media import (MediaRelay) from aiortc.contrib.media import (MediaRelay)
from av import AudioFifo from av import AudioFifo
@@ -18,13 +18,10 @@ from sortedcontainers import SortedDict
from whisper_jax import FlaxWhisperPipline from whisper_jax import FlaxWhisperPipline
from utils.log_utils import logger from utils.log_utils import logger
from utils.server_utils import Mutex from utils.run_utils import Mutex
ROOT = os.path.dirname(__file__) ROOT = os.path.dirname(__file__)
config = configparser.ConfigParser()
config.read('config.ini')
WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"] WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"]
pcs = set() pcs = set()
relay = MediaRelay() relay = MediaRelay()
@@ -91,10 +88,10 @@ def get_transcription():
wf.close() wf.close()
whisper_result = pipeline(out_file.getvalue()) whisper_result = pipeline(out_file.getvalue())
item = {'text': whisper_result["text"], item = { 'text': whisper_result["text"],
'start_time': str(frames[0].time), 'start_time': str(frames[0].time),
'time': str(datetime.datetime.now()) 'time': str(datetime.datetime.now())
} }
sorted_message_queue[frames[0].time] = str(item) sorted_message_queue[frames[0].time] = str(item)
start_messaging_thread() start_messaging_thread()
except Exception as e: except Exception as e:
@@ -177,10 +174,10 @@ async def offer(request):
answer = await pc.createAnswer() answer = await pc.createAnswer()
await pc.setLocalDescription(answer) await pc.setLocalDescription(answer)
return web.Response( return web.Response(
content_type="application/json", content_type="application/json",
text=json.dumps( text=json.dumps(
{"sdp": pc.localDescription.sdp, "type": pc.localDescription.type} { "sdp": pc.localDescription.sdp, "type": pc.localDescription.type }
), ),
) )
@@ -196,5 +193,5 @@ if __name__ == "__main__":
start_transcription_thread(6) start_transcription_thread(6)
app.router.add_post("/offer", offer) app.router.add_post("/offer", offer)
web.run_app( web.run_app(
app, access_log=None, host="127.0.0.1", port=1250 app, access_log=None, host="127.0.0.1", port=1250
) )

12
stream_client.py
View File

@@ -1,6 +1,6 @@
import ast import ast
import asyncio import asyncio
import configparser from utils.run_utils import config
import time import time
import uuid import uuid
@@ -12,12 +12,10 @@ from aiortc import (RTCPeerConnection, RTCSessionDescription)
from aiortc.contrib.media import (MediaPlayer, MediaRelay) from aiortc.contrib.media import (MediaPlayer, MediaRelay)
from utils.log_utils import logger from utils.log_utils import logger
from utils.server_utils import Mutex from utils.run_utils import Mutex
file_lock = Mutex(open("test_sm_6.txt", "a")) file_lock = Mutex(open("test_sm_6.txt", "a"))
config = configparser.ConfigParser()
config.read('config.ini')
class StreamClient: class StreamClient:
@@ -42,7 +40,7 @@ class StreamClient:
self.time_start = None self.time_start = None
self.queue = asyncio.Queue() self.queue = asyncio.Queue()
self.player = MediaPlayer(':' + str(config['DEFAULT']["AV_FOUNDATION_DEVICE_ID"]), self.player = MediaPlayer(':' + str(config['DEFAULT']["AV_FOUNDATION_DEVICE_ID"]),
format='avfoundation', options={'channels': '2'}) format='avfoundation', options={ 'channels': '2' })
def stop(self): def stop(self):
self.loop.run_until_complete(self.signaling.close()) self.loop.run_until_complete(self.signaling.close())
@@ -127,8 +125,8 @@ class StreamClient:
await pc.setLocalDescription(await pc.createOffer()) await pc.setLocalDescription(await pc.createOffer())
sdp = { sdp = {
"sdp": pc.localDescription.sdp, "sdp": pc.localDescription.sdp,
"type": pc.localDescription.type "type": pc.localDescription.type
} }
@stamina.retry(on=httpx.HTTPError, attempts=5) @stamina.retry(on=httpx.HTTPError, attempts=5)

18
utils/file_utils.py
View File

@@ -1,14 +1,12 @@
import configparser import configparser
import sys
import boto3 import boto3
import botocore import botocore
from run_utils import config
from log_utils import logger from log_utils import logger
config = configparser.ConfigParser() BUCKET_NAME = config["DEFAULT"]["BUCKET_NAME"]
config.read('config.ini')
BUCKET_NAME = 'reflector-bucket'
s3 = boto3.client('s3', s3 = boto3.client('s3',
aws_access_key_id=config["DEFAULT"]["AWS_ACCESS_KEY"], aws_access_key_id=config["DEFAULT"]["AWS_ACCESS_KEY"],
@@ -18,8 +16,8 @@ s3 = boto3.client('s3',
def upload_files(files_to_upload): def upload_files(files_to_upload):
""" """
Upload a list of files to the configured S3 bucket Upload a list of files to the configured S3 bucket
:param files_to_upload: :param files_to_upload: List of files to upload
:return: :return: None
""" """
for KEY in files_to_upload: for KEY in files_to_upload:
logger.info("Uploading file " + KEY) logger.info("Uploading file " + KEY)
@@ -32,8 +30,8 @@ def upload_files(files_to_upload):
def download_files(files_to_download): def download_files(files_to_download):
""" """
Download a list of files from the configured S3 bucket Download a list of files from the configured S3 bucket
:param files_to_download: :param files_to_download: List of files to download
:return: :return: None
""" """
for KEY in files_to_download: for KEY in files_to_download:
logger.info("Downloading file " + KEY) logger.info("Downloading file " + KEY)
@@ -47,8 +45,6 @@ def download_files(files_to_download):
if __name__ == "__main__": if __name__ == "__main__":
import sys
if sys.argv[1] == "download": if sys.argv[1] == "download":
download_files([sys.argv[2]]) download_files([sys.argv[2]])
elif sys.argv[1] == "upload": elif sys.argv[1] == "upload":

4
utils/log_utils.py
View File

@@ -6,6 +6,10 @@ class SingletonLogger:
@staticmethod @staticmethod
def get_logger(): def get_logger():
"""
Create or return the singleton instance for the SingletonLogger class
:return: SingletonLogger instance
"""
if not SingletonLogger.__instance: if not SingletonLogger.__instance:
SingletonLogger.__instance = logger SingletonLogger.__instance = logger
return SingletonLogger.__instance return SingletonLogger.__instance

66
utils/run_utils.py Normal file
View File

@@ -0,0 +1,66 @@
import asyncio
import configparser
import contextlib
from functools import partial
from threading import Lock
from typing import ContextManager, Generic, TypeVar
class ConfigParser:
__config = configparser.ConfigParser()
def __init__(self, config_file='../config.ini'):
self.__config.read(config_file)
@staticmethod
def get_config():
return ConfigParser.__config
config = ConfigParser.get_config()
def run_in_executor(func, *args, executor=None, **kwargs):
"""
Run the function in an executor, unblocking the main loop
:param func: Function to be run in executor
:param args: function parameters
:param executor: executor instance [Thread | Process]
:param kwargs: Additional parameters
:return: Future of function result upon completion
"""
callback = partial(func, *args, **kwargs)
loop = asyncio.get_event_loop()
return asyncio.get_event_loop().run_in_executor(executor, callback)
# Genetic type template
T = TypeVar("T")
class Mutex(Generic[T]):
"""
Mutex class to implement lock/release of a shared
protected variable
"""
def __init__(self, value: T):
"""
Create an instance of Mutex wrapper for the given resource
:param value: Shared resources to be thread protected
"""
self.__value = value
self.__lock = Lock()
@contextlib.contextmanager
def lock(self) -> ContextManager[T]:
"""
Lock the resource with a mutex to be used within a context block
The lock is automatically released on context exit
:return: Shared resource
"""
self.__lock.acquire()
try:
yield self.__value
finally:
self.__lock.release()

28
utils/server_utils.py
View File

@@ -1,28 +0,0 @@
import asyncio
import contextlib
from functools import partial
from threading import Lock
from typing import ContextManager, Generic, TypeVar
def run_in_executor(func, *args, executor=None, **kwargs):
callback = partial(func, *args, **kwargs)
loop = asyncio.get_event_loop()
return asyncio.get_event_loop().run_in_executor(executor, callback)
T = TypeVar("T")
class Mutex(Generic[T]):
def __init__(self, value: T):
self.__value = value
self.__lock = Lock()
@contextlib.contextmanager
def lock(self) -> ContextManager[T]:
self.__lock.acquire()
try:
yield self.__value
finally:
self.__lock.release()

8
utils/text_utilities.py
View File

@@ -6,14 +6,12 @@ from nltk.corpus import stopwords
from nltk.tokenize import word_tokenize from nltk.tokenize import word_tokenize
from sklearn.feature_extraction.text import TfidfVectorizer from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity from sklearn.metrics.pairwise import cosine_similarity
from transformers import BartTokenizer, BartForConditionalGeneration from transformers import BartForConditionalGeneration, BartTokenizer
from run_utils import config
from log_utils import logger from log_utils import logger
nltk.download('punkt', quiet=True) nltk.download('punkt', quiet=True)
config = configparser.ConfigParser()
config.read('config.ini')
def preprocess_sentence(sentence): def preprocess_sentence(sentence):
@@ -74,7 +72,7 @@ def remove_whisper_repetitive_hallucination(nonduplicate_sentences):
for sent in nonduplicate_sentences: for sent in nonduplicate_sentences:
temp_result = "" temp_result = ""
seen = {} seen = { }
words = nltk.word_tokenize(sent) words = nltk.word_tokenize(sent)
n_gram_filter = 3 n_gram_filter = 3
for i in range(len(words)): for i in range(len(words)):

28
utils/viz_utilities.py
View File

@@ -1,6 +1,5 @@
import ast import ast
import collections import collections
import configparser
import os import os
import pickle import pickle
from pathlib import Path from pathlib import Path
@@ -10,10 +9,7 @@ import pandas as pd
import scattertext as st import scattertext as st
import spacy import spacy
from nltk.corpus import stopwords from nltk.corpus import stopwords
from wordcloud import WordCloud, STOPWORDS from wordcloud import STOPWORDS, WordCloud
config = configparser.ConfigParser()
config.read('config.ini')
en = spacy.load('en_core_web_md') en = spacy.load('en_core_web_md')
spacy_stopwords = en.Defaults.stop_words spacy_stopwords = en.Defaults.stop_words
@@ -92,11 +88,11 @@ def create_talk_diff_scatter_viz(timestamp, real_time=False):
# create df for processing # create df for processing
df = pd.DataFrame.from_dict(res["chunks"]) df = pd.DataFrame.from_dict(res["chunks"])
covered_items = {} covered_items = { }
# ts: timestamp # ts: timestamp
# Map each timestamped chunk with top1 and top2 matched agenda # Map each timestamped chunk with top1 and top2 matched agenda
ts_to_topic_mapping_top_1 = {} ts_to_topic_mapping_top_1 = { }
ts_to_topic_mapping_top_2 = {} ts_to_topic_mapping_top_2 = { }
# Also create a mapping of the different timestamps in which each topic was covered # Also create a mapping of the different timestamps in which each topic was covered
topic_to_ts_mapping_top_1 = collections.defaultdict(list) topic_to_ts_mapping_top_1 = collections.defaultdict(list)
@@ -189,16 +185,16 @@ def create_talk_diff_scatter_viz(timestamp, real_time=False):
# Scatter plot of topics # Scatter plot of topics
df = df.assign(parse=lambda df: df.text.apply(st.whitespace_nlp_with_sentences)) df = df.assign(parse=lambda df: df.text.apply(st.whitespace_nlp_with_sentences))
corpus = st.CorpusFromParsedDocuments( corpus = st.CorpusFromParsedDocuments(
df, category_col='ts_to_topic_mapping_top_1', parsed_col='parse' df, category_col='ts_to_topic_mapping_top_1', parsed_col='parse'
).build().get_unigram_corpus().compact(st.AssociationCompactor(2000)) ).build().get_unigram_corpus().compact(st.AssociationCompactor(2000))
html = st.produce_scattertext_explorer( html = st.produce_scattertext_explorer(
corpus, corpus,
category=cat_1, category=cat_1,
category_name=cat_1_name, category_name=cat_1_name,
not_category_name=cat_2_name, not_category_name=cat_2_name,
minimum_term_frequency=0, pmi_threshold_coefficient=0, minimum_term_frequency=0, pmi_threshold_coefficient=0,
width_in_pixels=1000, width_in_pixels=1000,
transform=st.Scalers.dense_rank transform=st.Scalers.dense_rank
) )
if real_time: if real_time:
open('./artefacts/real_time_scatter_' + timestamp.strftime("%m-%d-%Y_%H:%M:%S") + '.html', 'w').write(html) open('./artefacts/real_time_scatter_' + timestamp.strftime("%m-%d-%Y_%H:%M:%S") + '.html', 'w').write(html)

29
whisjax.py
View File

@@ -20,18 +20,15 @@ import nltk
import yt_dlp as youtube_dl import yt_dlp as youtube_dl
from whisper_jax import FlaxWhisperPipline from whisper_jax import FlaxWhisperPipline
from utils.file_utils import upload_files, download_files from utils.file_utils import download_files, upload_files
from utils.log_utils import logger from utils.log_utils import logger
from utils.text_utilities import summarize, post_process_transcription from utils.run_utils import config
from utils.viz_utilities import create_wordcloud, create_talk_diff_scatter_viz from utils.text_utilities import post_process_transcription, summarize
from utils.viz_utilities import create_talk_diff_scatter_viz, create_wordcloud
nltk.download('punkt', quiet=True) nltk.download('punkt', quiet=True)
nltk.download('stopwords', quiet=True) nltk.download('stopwords', quiet=True)
# Configurations can be found in config.ini. Set them properly before executing
config = configparser.ConfigParser()
config.read('config.ini')
WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"] WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"]
NOW = datetime.now() NOW = datetime.now()
@@ -42,8 +39,8 @@ def init_argparse() -> argparse.ArgumentParser:
:return: parser object :return: parser object
""" """
parser = argparse.ArgumentParser( parser = argparse.ArgumentParser(
usage="%(prog)s [OPTIONS] <LOCATION> <OUTPUT>", usage="%(prog)s [OPTIONS] <LOCATION> <OUTPUT>",
description="Creates a transcript of a video or audio file, then summarizes it using ChatGPT." description="Creates a transcript of a video or audio file, then summarizes it using ChatGPT."
) )
parser.add_argument("-l", "--language", help="Language that the summary should be written in", type=str, parser.add_argument("-l", "--language", help="Language that the summary should be written in", type=str,
@@ -74,13 +71,13 @@ def main():
# Create options for the download # Create options for the download
ydl_opts = { ydl_opts = {
'format': 'bestaudio/best', 'format': 'bestaudio/best',
'postprocessors': [{ 'postprocessors': [{
'key': 'FFmpegExtractAudio', 'key': 'FFmpegExtractAudio',
'preferredcodec': 'mp3', 'preferredcodec': 'mp3',
'preferredquality': '192', 'preferredquality': '192',
}], }],
'outtmpl': 'audio', # Specify the output file path and name 'outtmpl': 'audio', # Specify the output file path and name
} }
# Download the audio # Download the audio

29
whisjax_realtime.py
View File

@@ -13,11 +13,10 @@ from whisper_jax import FlaxWhisperPipline
from utils.file_utils import upload_files from utils.file_utils import upload_files
from utils.log_utils import logger from utils.log_utils import logger
from utils.text_utilities import summarize, post_process_transcription from utils.run_utils import config
from utils.viz_utilities import create_wordcloud, create_talk_diff_scatter_viz from utils.text_utilities import post_process_transcription, summarize
from utils.viz_utilities import create_talk_diff_scatter_viz, create_wordcloud
config = configparser.ConfigParser()
config.read('config.ini')
WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"] WHISPER_MODEL_SIZE = config['DEFAULT']["WHISPER_MODEL_SIZE"]
@@ -37,12 +36,12 @@ def main():
AUDIO_DEVICE_ID = i AUDIO_DEVICE_ID = i
audio_devices = p.get_device_info_by_index(AUDIO_DEVICE_ID) audio_devices = p.get_device_info_by_index(AUDIO_DEVICE_ID)
stream = p.open( stream = p.open(
format=FORMAT, format=FORMAT,
channels=CHANNELS, channels=CHANNELS,
rate=RATE, rate=RATE,
input=True, input=True,
frames_per_buffer=FRAMES_PER_BUFFER, frames_per_buffer=FRAMES_PER_BUFFER,
input_device_index=int(audio_devices['index']) input_device_index=int(audio_devices['index'])
) )
pipeline = FlaxWhisperPipline("openai/whisper-" + config["DEFAULT"]["WHISPER_REAL_TIME_MODEL_SIZE"], pipeline = FlaxWhisperPipline("openai/whisper-" + config["DEFAULT"]["WHISPER_REAL_TIME_MODEL_SIZE"],
@@ -60,7 +59,7 @@ def main():
global proceed global proceed
proceed = False proceed = False
transcript_with_timestamp = {"text": "", "chunks": []} transcript_with_timestamp = { "text": "", "chunks": [] }
last_transcribed_time = 0.0 last_transcribed_time = 0.0
listener = keyboard.Listener(on_press=on_press) listener = keyboard.Listener(on_press=on_press)
@@ -90,10 +89,10 @@ def main():
if end is None: if end is None:
end = start + 15.0 end = start + 15.0
duration = end - start duration = end - start
item = {'timestamp': (last_transcribed_time, last_transcribed_time + duration), item = { 'timestamp': (last_transcribed_time, last_transcribed_time + duration),
'text': whisper_result['text'], 'text': whisper_result['text'],
'stats': (str(end_time - start_time), str(duration)) 'stats': (str(end_time - start_time), str(duration))
} }
last_transcribed_time = last_transcribed_time + duration last_transcribed_time = last_transcribed_time + duration
transcript_with_timestamp["chunks"].append(item) transcript_with_timestamp["chunks"].append(item)
transcription += whisper_result['text'] transcription += whisper_result['text']