Skip to content

Commit

Permalink
Refactor + toggle switch
Browse files Browse the repository at this point in the history
  • Loading branch information
@rmusser01
rmusser01 committed May 14, 2024
1 parent c6dec4b commit 025041d
Show file tree
Hide file tree
Showing 3 changed files with 109 additions and 196 deletions.
174 changes: 48 additions & 126 deletions HF/app.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -349,40 +349,31 @@ def process_local_file(file_path):
# Video Download/Handling
#

def process_url(input_path, num_speakers=2, whisper_model="small.en", custom_prompt=None, offset=0, api_name=None,
api_key=None, vad_filter=False, download_video_flag=False, demo_mode=False):
if demo_mode:
api_name = "huggingface"
api_key = os.environ.get(HF_TOKEN)
print("HUGGINGFACE API KEY CHECK #3: " + api_key)
vad_filter = False
download_video_flag = False

def process_url(url, num_speakers, whisper_model, custom_prompt, offset, api_name, api_key, vad_filter,
download_video, download_audio):
video_file_path = None
try:
results = main(input_path, api_name=api_name, api_key=api_key, num_speakers=num_speakers,
results = main(url, api_name=api_name, api_key=api_key, num_speakers=num_speakers,
whisper_model=whisper_model, offset=offset, vad_filter=vad_filter,
download_video_flag=download_video_flag)

download_video_flag=download_video, custom_prompt=custom_prompt)
if results:
transcription_result = results[0]
json_file_path = transcription_result['audio_file'].replace('.wav', '.segments.json')

with open(json_file_path, 'r') as file:
json_data = json.load(file)
summary_file_path = json_file_path.replace('.segments.json', '_summary.txt')

if os.path.exists(summary_file_path):
return json_data, summary_file_path, json_file_path, summary_file_path
json_file_path = format_file_path(json_file_path)
summary_file_path = format_file_path(summary_file_path)

if summary_file_path and os.path.exists(summary_file_path):
return transcription_result[
'transcription'], "Summary available", json_file_path, summary_file_path, video_file_path
else:
return json_data, "Summary not available.", json_file_path, "Summary not available."

return transcription_result[
'transcription'], "Summary not available", json_file_path, None, video_file_path
else:
return None, "No results found.", None, None

return "No results found.", "Summary not available", None, None, None
except Exception as e:
error_message = f"An error occurred: {str(e)}"
return None, error_message, None, None
return str(e), "Error processing the request.", None, None, None


def create_download_directory(title):
Expand Down Expand Up @@ -1244,125 +1235,56 @@ def format_file_path(file_path):

def launch_ui(demo_mode=False):
inputs = [
gr.components.Textbox(label="URL", placeholder="Enter the video URL here"),
gr.components.Number(value=2, label="Number of Speakers"),
gr.components.Dropdown(choices=whisper_models, value="small.en", label="Whisper Model"),
gr.components.Textbox(label="Custom Prompt", placeholder="Q: As a professional summarizer, create a concise and comprehensive summary of the provided text.\nA: Here is a detailed, bulleted list of the key points made in the transcribed video and supporting arguments:", lines=3),
gr.components.Number(value=0, label="Offset"),
gr.components.Dropdown(
choices=["huggingface", "openai", "anthropic", "cohere", "groq", "llama", "kobold", "ooba"],
label="API Name"),
gr.components.Textbox(label="API Key", placeholder="Enter your API key here"),
gr.components.Checkbox(label="VAD Filter", value=False),
gr.components.Checkbox(label="Download Video", value=False)
]

outputs = [
gr.components.Textbox(label="Transcription"),
gr.components.Textbox(label="Summary or Status Message"),
gr.components.File(label="Download Transcription as JSON", visible=lambda x: x != "File not available"),
gr.components.File(label="Download Summary as Text", visible=lambda x: x != "File not available"),
gr.components.File(label="Download Video", visible=lambda x: x is not None)
]

def process_url(url, num_speakers, whisper_model, custom_prompt, offset, api_name, api_key, vad_filter,
download_video):
video_file_path = None
try:
results = main(url, api_name=api_name, api_key=api_key, num_speakers=num_speakers,
whisper_model=whisper_model, offset=offset, vad_filter=vad_filter,
download_video_flag=download_video, custom_prompt=custom_prompt)
if results:
transcription_result = results[0]
json_file_path = transcription_result['audio_file'].replace('.wav', '.segments.json')
summary_file_path = json_file_path.replace('.segments.json', '_summary.txt')

json_file_path = format_file_path(json_file_path)
summary_file_path = format_file_path(summary_file_path)

if summary_file_path and os.path.exists(summary_file_path):
return transcription_result[
'transcription'], "Summary available", json_file_path, summary_file_path, video_file_path
else:
return transcription_result[
'transcription'], "Summary not available", json_file_path, None, video_file_path
else:
return "No results found.", "Summary not available", None, None, None
except Exception as e:
return str(e), "Error processing the request.", None, None, None

iface = gr.Interface(
fn=process_url,
inputs=inputs,
outputs=outputs,
title="Video Transcription and Summarization",
description="Submit a video URL for transcription and summarization. Ensure you input all necessary information including API keys."
)

iface.launch(share=False)




a = """def launch_ui(demo_mode=False):
def process_url(url, num_speakers, whisper_model, custom_prompt, offset, api_name, api_key, vad_filter,
download_video):
try:
results = main(url, api_name=api_name, api_key=api_key, num_speakers=num_speakers,
whisper_model=whisper_model, offset=offset, vad_filter=vad_filter,
download_video_flag=download_video, custom_prompt=custom_prompt)
if results:
transcription_result = results[0]
json_data = transcription_result['transcription']
json_file_path = transcription_result['audio_file'].replace('.wav', '.segments.json')
summary_file_path = transcription_result.get('summary', "Summary not available.")
video_file_path = transcription_result.get('video_path', None)
json_file_path = format_file_path(json_file_path)
summary_file_path = format_file_path(summary_file_path)
return json_data, "Summary available", json_file_path, summary_file_path, video_file_path
else:
return "No results found.", "No summary available.", None, None, None
except Exception as e:
return str(e), "Error processing the request.", None, None, None, None
inputs = [
gr.components.Textbox(label="URL", placeholder="Enter the video URL here"),
gr.components.Number(value=2, label="Number of Speakers"),
gr.components.Dropdown(choices=whisper_models, value="small.en", label="Whisper Model"),
gr.components.Textbox(label="Custom Prompt",
placeholder="Q: As a professional summarizer, create a concise and comprehensive summary of the provided text.\nA: Here is a detailed, bulleted list of the key points made in the transcribed video and supporting arguments:",
gr.components.Textbox(label="URL (Mandatory)",
placeholder="Enter the video URL here"),
gr.components.Number(value=2,
label="Number of Speakers(Optional - Currently has no effect)",
visible=lambda x: x is not None),
gr.components.Dropdown(choices=whisper_models,
value="small.en",
label="Whisper Model(This is the ML model used for transcription.)",
visible=lambda x: x is not None),
gr.components.Textbox(label="Custom Prompt (Customize your summary, or ask a different question)",
placeholder="Q: As a professional summarizer, create a concise and comprehensive "
"summary of the provided text.\nA: Here is a detailed, bulleted list of the "
"key points made in the transcribed video and supporting arguments:",
lines=3),
gr.components.Number(value=0, label="Offset"),
gr.components.Number(value=0,
label="Offset (Seconds into the video to start transcribing at)",
visible=lambda x: x is not None),
gr.components.Dropdown(
choices=["huggingface", "openai", "anthropic", "cohere", "groq", "llama", "kobold", "ooba"],
label="API Name"),
gr.components.Textbox(label="API Key", placeholder="Enter your API key here"),
gr.components.Checkbox(label="VAD Filter", value=False),
gr.components.Checkbox(label="Download Video", value=False)
label="API Name (Mandatory Unless you just want a Transcription)", visible=lambda x: x is not None),
gr.components.Textbox(label="API Key (Mandatory if API Name is specified)",
placeholder="Enter your API key here", visible=lambda x: x is not None),
gr.components.Checkbox(label="VAD Filter(Can safely ignore)",
value=False, visible=lambda x: x is not None),
gr.components.Checkbox(label="Download Video(Select to allow for file download of selected video)",
value=False, visible=lambda x: x is not None),
gr.components.Checkbox(label="Download Audio(Select to allow for file download of selected Video's Audio)",
value=False, visible=lambda x: x is not None)
]

outputs = [
gr.components.Textbox(label="Transcription"),
gr.components.Textbox(label="Summary or Status Message"),
gr.components.File(label="Download Transcription as JSON", visible=lambda x: x != "File not available"),
gr.components.File(label="Download Summary as Text", visible=lambda x: x != "File not available"),
gr.components.File(label="Download Video", visible=lambda x: x is not None)
gr.components.Textbox(label="Transcription (Resulting Transcription from your input URL)"),
gr.components.Textbox(label="Summary or Status Message (Current status of Summary or Summary itself)"),
gr.components.File(label="Download Transcription as JSON (Download the Transcription as a file)",
visible=lambda x: x is not None),
gr.components.File(label="Download Summary as Text (Download the Summary as a file)",
visible=lambda x: x is not None),
gr.components.File(label="Download Video (Download the Video as a file)", visible=lambda x: x is not None),
gr.components.File(label="Download Audio (Download the Audio as a file)", visible=lambda x: x is not None)
]

iface = gr.Interface(
fn=process_url,
inputs=inputs,
outputs=outputs,
title="Video Transcription and Summarization",
description="Submit a video URL for transcription and summarization. Ensure you input all necessary information including API keys.",
theme="bethecloud/storj_theme" # Adjust theme as necessary
description="Submit a video URL for transcription and summarization. Ensure you input all necessary information including API keys."
)

iface.launch(share=False)
"""

#
#
Expand Down
3 changes: 3 additions & 0 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -242,6 +242,9 @@ By default videos, transcriptions and summaries are stored in a folder with the
- `compare.py` - prepare LLM outputs for webapp
- `compare-app.py` - summary viewer webapp

------------
### Similar/Other projects:
- https://github.com/Dicklesworthstone/bulk_transcribe_youtube_videos_from_playlist/tree/main

------------

Expand Down
Loading

0 comments on commit 025041d

Please sign in to comment.