scripts

Author: spech66

dalle.py

@@ -0,0 +1,46 @@
+import os
+import sys
+from datetime import datetime
+import urllib.request
+from dotenv import load_dotenv
+import openai
+from openai import OpenAI
+
+load_dotenv() 
+
+# Fail if no commandline argument is provided
+if len(sys.argv) < 2:
+    print("Please provide an text")
+    exit(1)
+
+text_prompt = sys.argv[1]
+
+client = OpenAI()
+
+image_url = ""
+
+try:
+    response = client.images.generate(
+    model="dall-e-3",
+    prompt=text_prompt, # "a white siamese cat",
+    size="1024x1024",
+    quality="standard",
+    n=1,
+    )
+    image_url = response.data[0].url
+except openai.OpenAIError as e:
+  print(e.http_status)
+  print(e.error)
+  exit(1)
+
+print(image_url)
+
+# Download image using python and save it to file in the dalle folder
+if not os.path.isdir("dalle"):
+    os.mkdir("dalle")
+with urllib.request.urlopen(image_url) as response:
+    cur_time = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
+    with open(f"dalle/{cur_time}.png", "wb") as f:
+        f.write(response.read())
+    with open(f"dalle/{cur_time}.txt", "wb") as f:
+        f.write(text_prompt.encode("utf-8"))

voice.py

@@ -0,0 +1,43 @@
+import os
+import sys
+from datetime import datetime
+import urllib.request
+from dotenv import load_dotenv
+import openai
+from openai import OpenAI
+from pathlib import Path
+
+load_dotenv() 
+
+# Fail if no commandline argument is provided
+if len(sys.argv) < 2:
+    print("Please provide an text")
+    exit(1)
+
+text_prompt = sys.argv[1]
+
+client = OpenAI()
+
+speech_file_path = ""
+
+# Stream audio and save it to file in the voice folder
+if not os.path.isdir("voice"):
+    os.mkdir("voice")
+    
+cur_time = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
+
+try:
+    speech_file_path = f"voice/{cur_time}.mp3"
+    response = client.audio.speech.create(
+        model="tts-1",
+        voice="alloy",
+        input=text_prompt
+    )
+
+    response.stream_to_file(speech_file_path)
+except openai.OpenAIError as e:
+  print(e.http_status)
+  print(e.error)
+  exit(1)
+
+print(speech_file_path)

whisper-yt.py

@@ -0,0 +1,42 @@
+import os
+import sys
+import yt_dlp
+import openai
+
+# Fail if no commandline argument is provided
+if len(sys.argv) < 2:
+    print("Please provide an video link")
+    exit(1)
+
+video_file_url = sys.argv[1]
+
+# https://github.com/yt-dlp/yt-dlp#filter-videos
+ydl_opts = {
+    'format': 'm4a/bestaudio/best',
+    'postprocessors': [{  # Extract audio using ffmpeg
+        'key': 'FFmpegExtractAudio',
+        'preferredcodec': 'm4a',
+    }]
+}
+
+audio_file_name = ""
+with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+    info = ydl.extract_info(video_file_url, download=False)
+    audio_file_name = ydl.prepare_filename(info)
+    ydl.process_info(info)  # starts the download
+
+# Check if file exists
+if not os.path.isfile(audio_file_name):
+    print("File {audio_file_name} does not exist")
+    exit(1)
+
+audio_file = open(audio_file_name, "rb")
+transcript = openai.Audio.transcribe("whisper-1", audio_file)
+
+print(transcript)
+
+# write transcript to file
+if not os.path.isdir("whisper_yt"):
+    os.mkdir("whisper_yt")
+with open(f"whisper_yt/{audio_file_name}.txt", "w") as f:
+    f.write(transcript["text"])

whisper.py

@@ -0,0 +1,30 @@
+import os
+import sys
+import openai
+
+# Fail if no commandline argument is provided
+if len(sys.argv) < 2:
+    print("Please provide an audio file")
+    exit(1)
+
+audio_file_name = sys.argv[1]
+
+# Check if file exists
+if not os.path.isfile(audio_file_name):
+    print("File does not exist")
+    exit(1)
+
+# If file is in ogg format convert it to mp3
+if audio_file_name.endswith(".ogg"):
+    os.system(f"ffmpeg -i '{audio_file_name}' -ab 320k '{audio_file_name}.mp3'")
+    audio_file_name = f"{audio_file_name}.mp3"
+
+audio_file= open(audio_file_name, "rb")
+transcript = openai.Audio.transcribe("whisper-1", audio_file, language="de")
+print(transcript)
+
+# write transcript to file
+if not os.path.isdir("whisper_audio"):
+    os.mkdir("whisper_audio")
+with open(f"whisper_audio/{audio_file_name}.txt", "w") as f:
+    f.write(transcript["text"])