add youtube video transcription example

Author: eff-kay

.gitignore

@@ -1,3 +1,4 @@
+examples/*/data/*
 __pycache__
 *.egg-info/
 

…ormer/deploy_hugging_face_transformer.py …ormer/deploy_hugging_face_transformer.pyexamples/deploy_model/hugging_face_transformer/deploy_hugging_face_transformer.py renamed to examples/deploy_custom_models/hugging_face_transformer/deploy_hugging_face_transformer.py examples/deploy_model/hugging_face_transformer/deploy_hugging_face_transformer.py renamed to examples/deploy_custom_models/hugging_face_transformer/deploy_hugging_face_transformer.py

@@ -0,0 +1 @@
+youtube-dl@git+https://github.com/ytdl-org/youtube-dl.git

…ugging_face_transformer/requirements.txt …ugging_face_transformer/requirements.txtexamples/deploy_model/hugging_face_transformer/requirements.txt renamed to examples/deploy_custom_models/hugging_face_transformer/requirements.txt examples/deploy_model/hugging_face_transformer/requirements.txt renamed to examples/deploy_custom_models/hugging_face_transformer/requirements.txt

@@ -0,0 +1,40 @@
+
+# import the model
+from slashml import SpeechToText
+from utils import download_youtube_file
+
+# Replace `API_KEY` with your SlasML API token. This example still runs without
+# the API token but usage will be limited
+API_KEY = None
+
+def transcribe_youtube_file(*, youtube_url:str):
+
+    # parse the id from the youtube url
+    youtube_id = youtube_url.split('/')[-1]
+
+    # this downloads the file at data/{youtube_id}.mp4 
+    download_youtube_file(youtube_url = youtube_url, output_folder='data')
+    
+    downloaded_file_path= f'data/{youtube_id}.mp4'
+
+    # We recommend using AWS for video transcription
+    service_provider = SpeechToText.ServiceProvider.AWS
+
+    model = SpeechToText(api_key=API_KEY)
+
+    # Upload audio
+    uploaded_file = model.upload_audio(downloaded_file_path)
+    print(f"file uploaded: {uploaded_file}")
+
+    response = model.execute(
+        upload_url=uploaded_file["upload_url"], service_provider=service_provider
+    )
+
+    return response.transcription_data.transcription
+
+def transcribe_multiple_youtube_videos(*, youtube_urls:list):
+    return list(map(transcribe_youtube_file, youtube_urls))
+
+if __name__=='__main__':
+    youtube_url = 'https://youtu.be/5-TgqZ8nado'
+    print(transcribe_youtube_file(youtube_url=youtube_url))

…y_model/pytorch/deploy_pytorch_models.py …_models/pytorch/deploy_pytorch_models.pyexamples/deploy_model/pytorch/deploy_pytorch_models.py renamed to examples/deploy_custom_models/pytorch/deploy_pytorch_models.py examples/deploy_model/pytorch/deploy_pytorch_models.py renamed to examples/deploy_custom_models/pytorch/deploy_pytorch_models.py

@@ -0,0 +1,10 @@
+import youtube_dl
+
+def download_youtube_file(*, youtube_url:str, output_folder:str):
+    ydl = youtube_dl.YoutubeDL({'outtmpl': output_folder+'/%(id)s.%(ext)s'})
+
+    with ydl:
+        ydl.download([youtube_url])
+    
+if __name__=='__main__':
+    download_youtube_file(youtube_url='https://youtu.be/5-TgqZ8nado', output_folder='data')