Skip to content

Commit 983ba12

Browse files
Jon BennettJon Bennett
Jon Bennett
authored and
Jon Bennett
committed
5/4/24-8:38 AM
0 parents  commit 983ba12

File tree

64 files changed

+2265
-0
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

64 files changed

+2265
-0
lines changed

‎.DS_Store

6 KB
Binary file not shown.

‎.gitignore

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,2 @@
1+
# .gitignore content
2+
.env

‎11_labs copy.py

Lines changed: 120 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,120 @@
1+
import os
2+
import openai
3+
import requests
4+
import boto3
5+
from mutagen.mp3 import MP3
6+
from dotenv import load_dotenv
7+
8+
# Setup basic configuration for logging
9+
import logging
10+
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
11+
12+
# Load environment variables
13+
load_dotenv()
14+
15+
# Constants
16+
OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')
17+
XI_API_KEY = os.getenv('XI_API_KEY')
18+
BUCKET_NAME = "cashdaily1"
19+
VOICE_ID = "4dyF1bD2mjNHsB76Wf9q"
20+
21+
if not OPENAI_API_KEY or not XI_API_KEY:
22+
logging.error("API keys are not set in environment variables.")
23+
exit(1)
24+
25+
# Initialize OpenAI client
26+
openai.api_key = OPENAI_API_KEY
27+
28+
# Step 1. Ask user for the relative path to the script.txt file and the step to start from
29+
text_file_path = input("Enter the relative path to your script.txt file: ")
30+
start_step = int(input("Enter the step number to start from (1-10): "))
31+
32+
if not os.path.isfile(text_file_path):
33+
logging.error(f"The file {text_file_path} does not exist.")
34+
exit(1)
35+
36+
text_file_dir = os.path.dirname(os.path.abspath(text_file_path))
37+
38+
# Step 2. Read text from file
39+
if start_step <= 2:
40+
try:
41+
with open(text_file_path, 'r') as file:
42+
text_to_speak = file.read()
43+
logging.info("Text read successfully from script.txt")
44+
except Exception as e:
45+
logging.error(f"Failed to read text file: {str(e)}")
46+
exit(1)
47+
48+
# Step 3. Generate a folder name using OpenAI
49+
if start_step <= 3:
50+
try:
51+
response = openai.ChatCompletion.create(
52+
model="gpt-3.5-turbo",
53+
messages=[{"role": "system", "content": "Generate a concise, descriptive name for a folder based on the following text. Don't use special characters or spaces"},
54+
{"role": "user", "content": text_to_speak}],
55+
max_tokens=10
56+
)
57+
folder_name = response['choices'][0]['message']['content'].strip()
58+
folder_path = os.path.join(text_file_dir, folder_name)
59+
if not os.path.exists(folder_path):
60+
os.makedirs(folder_path)
61+
logging.info(f"Folder created: {folder_path}")
62+
except Exception as e:
63+
logging.error(f"OpenAI API error or folder creation error: {str(e)}")
64+
exit(1)
65+
66+
67+
# Step 4: Send text to elevenlabs for voice mp3 generation
68+
if start_step <= 6:
69+
tts_url = f"https://api.elevenlabs.io/v1/text-to-speech/{VOICE_ID}/stream"
70+
headers = {
71+
"Accept": "application/json",
72+
"xi-api-key": XI_API_KEY
73+
}
74+
data = {
75+
"text": text_to_speak,
76+
"model_id": "eleven_multilingual_v2",
77+
"voice_settings": {
78+
"stability": 0.5,
79+
"similarity_boost": 0.8,
80+
"style": 0.5,
81+
"use_speaker_boost": True
82+
}
83+
}
84+
response = requests.post(tts_url, headers=headers, json=data, stream=True)
85+
if response.ok:
86+
logging.info("Text-to-Speech API call successful")
87+
# Save the audio stream to a file
88+
LOCAL_FILE_PATH = os.path.join(folder_path, f"{folder_name}.mp3")
89+
with open(LOCAL_FILE_PATH, "wb") as f:
90+
for chunk in response.iter_content(chunk_size=1024):
91+
f.write(chunk)
92+
logging.info(f"Audio stream saved successfully as {LOCAL_FILE_PATH}")
93+
94+
# Step 4: Find length of mp3 file
95+
try:
96+
audio = MP3(LOCAL_FILE_PATH)
97+
total_length = int(audio.info.length)
98+
logging.info(f"Audio length: {total_length} seconds")
99+
100+
# Step 6: Calculate total scenes
101+
total_scenes = round(total_length / 5)
102+
logging.info(f"Total scenes calculated: {total_scenes}")
103+
104+
# Step 7: Upload mp3 file to s3
105+
S3_KEY = f"{folder_name}/{os.path.basename(LOCAL_FILE_PATH)}"
106+
s3_client = boto3.client('s3')
107+
s3_client.upload_file(LOCAL_FILE_PATH, BUCKET_NAME, S3_KEY)
108+
s3_url = f"https://{BUCKET_NAME}.s3.amazonaws.com/{S3_KEY}"
109+
logging.info(f"File uploaded to S3 at URL: {s3_url}")
110+
111+
# Step 10: Save the URL, audio length, and total scenes to assets.txt
112+
with open(os.path.join(folder_path, 'assets.txt'), 'a') as file:
113+
file.write(f"MP3 URL: {s3_url}\n")
114+
file.write(f"Total Length: {total_length} seconds\n")
115+
file.write(f"Total Scenes: {total_scenes}\n")
116+
logging.info("URL, audio length, and total scenes saved to assets.txt in the script directory")
117+
except Exception as e:
118+
logging.error(f"Error during MP3 handling or S3 upload: {str(e)}")
119+
else:
120+
logging.error(f"Failed to generate speech: {response.text}")

‎11_labs.py

Lines changed: 251 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,251 @@
1+
import os
2+
import openai
3+
import requests
4+
import boto3
5+
from mutagen.mp3 import MP3
6+
from dotenv import load_dotenv
7+
8+
# Setup basic configuration for logging
9+
import logging
10+
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
11+
12+
# Load environment variables
13+
load_dotenv()
14+
15+
# Constants
16+
OPENAI_API_KEY = os.getenv('OPENAI_API_KEY')
17+
XI_API_KEY = os.getenv('XI_API_KEY')
18+
BUCKET_NAME = "cashdaily1"
19+
VOICE_ID = "4dyF1bD2mjNHsB76Wf9q"
20+
21+
if not OPENAI_API_KEY or not XI_API_KEY:
22+
logging.error("API keys are not set in environment variables.")
23+
exit(1)
24+
25+
# Initialize OpenAI client
26+
openai.api_key = OPENAI_API_KEY
27+
28+
# Step 1. Ask user for the relative path to the script.txt file and the step to start from
29+
text_file_path = input("Enter the relative path to your script.txt file: ")
30+
start_step = int(input("Enter the step number to start from (1-10): "))
31+
32+
if not os.path.isfile(text_file_path):
33+
logging.error(f"The file {text_file_path} does not exist.")
34+
exit(1)
35+
36+
text_file_dir = os.path.dirname(os.path.abspath(text_file_path))
37+
38+
# Step 2. Read text from file
39+
if start_step <= 2:
40+
try:
41+
with open(text_file_path, 'r') as file:
42+
text_to_speak = file.read()
43+
logging.info("Text read successfully from script.txt")
44+
except Exception as e:
45+
logging.error(f"Failed to read text file: {str(e)}")
46+
exit(1)
47+
48+
# Step 3. Generate a folder name using OpenAI
49+
if start_step <= 3:
50+
try:
51+
response = openai.Completion.create(
52+
model="gpt-3.5-turbo",
53+
prompt=f"Generate a concise, descriptive name for a folder based on the following text: {text_to_speak}",
54+
max_tokens=10,
55+
temperature=0.7
56+
)
57+
folder_name = response.choices[0].text.strip()
58+
folder_path = os.path.join(text_file_dir, folder_name)
59+
if not os.path.exists(folder_path):
60+
os.makedirs(folder_path)
61+
logging.info(f"Folder created: {folder_path}")
62+
except Exception as e:
63+
logging.error(f"OpenAI API error or folder creation error: {str(e)}")
64+
exit(1)
65+
66+
67+
68+
# Step 4: Send text to elevenlabs for voice mp3 generation
69+
if start_step <= 6:
70+
tts_url = f"https://api.elevenlabs.io/v1/text-to-speech/{VOICE_ID}/stream"
71+
headers = {
72+
"Accept": "application/json",
73+
"xi-api-key": XI_API_KEY
74+
}
75+
data = {
76+
"text": text_to_speak,
77+
"model_id": "eleven_multilingual_v2",
78+
"voice_settings": {
79+
"stability": 0.5,
80+
"similarity_boost": 0.8,
81+
"style": 0.5,
82+
"use_speaker_boost": True
83+
}
84+
}
85+
response = requests.post(tts_url, headers=headers, json=data, stream=True)
86+
if response.ok:
87+
logging.info("Text-to-Speech API call successful")
88+
# Save the audio stream to a file
89+
LOCAL_FILE_PATH = os.path.join(folder_path, f"{folder_name}.mp3")
90+
with open(LOCAL_FILE_PATH, "wb") as f:
91+
for chunk in response.iter_content(chunk_size=1024):
92+
f.write(chunk)
93+
logging.info(f"Audio stream saved successfully as {LOCAL_FILE_PATH}")
94+
95+
# Step 4: Find length of mp3 file
96+
try:
97+
audio = MP3(LOCAL_FILE_PATH)
98+
total_length = int(audio.info.length)
99+
logging.info(f"Audio length: {total_length} seconds")
100+
101+
# Step 6: Calculate total scenes
102+
total_scenes = round(total_length / 5)
103+
logging.info(f"Total scenes calculated: {total_scenes}")
104+
105+
# Step 7: Upload mp3 file to s3
106+
S3_KEY = f"{folder_name}/{os.path.basename(LOCAL_FILE_PATH)}"
107+
s3_client = boto3.client('s3')
108+
s3_client.upload_file(LOCAL_FILE_PATH, BUCKET_NAME, S3_KEY)
109+
s3_url = f"https://{BUCKET_NAME}.s3.amazonaws.com/{S3_KEY}"
110+
logging.info(f"File uploaded to S3 at URL: {s3_url}")
111+
112+
# Step 8: Save the URL, audio length, and total scenes to assets.txt
113+
with open(os.path.join(folder_path, 'assets.txt'), 'a') as file:
114+
file.write(f"MP3 URL: {s3_url}\n")
115+
file.write(f"Total Length: {total_length} seconds\n")
116+
file.write(f"Total Scenes: {total_scenes}\n")
117+
logging.info("URL, audio length, and total scenes saved to assets.txt in the script directory")
118+
except Exception as e:
119+
logging.error(f"Error during MP3 handling or S3 upload: {str(e)}")
120+
else:
121+
logging.error(f"Failed to generate speech: {response.text}")
122+
123+
# Begin Image generation process. 1 image per scene
124+
# Define the API URL and headers
125+
post_url = "https://cloud.leonardo.ai/api/rest/v1/generations"
126+
api_key = os.getenv('LEONARDO_API_KEY')
127+
headers = {
128+
"accept": "application/json",
129+
"content-type": "application/json",
130+
"authorization": f"Bearer {api_key}"
131+
}
132+
133+
# Step 9: Improve prompts for images. Define prompt improvement URL
134+
prompt_improve_url = "https://cloud.leonardo.ai/api/rest/v1/prompt/improve"
135+
136+
# Predefined prompts
137+
prompts = [
138+
"a successful young female digital marketer in a luxurious setting (home, yacht, cafe, beach, poolside, or similar), using high-tech devices for her work",
139+
"a successful young female digital marketer in a luxurious setting enjoying her leisure time that success has afforded. Having a fun time with friends or family"
140+
]
141+
142+
# Step 10: Read the total scenes from assets.txt to determine the number of images to generate
143+
assets_file_path = os.path.join(folder_path, 'assets.txt')
144+
try:
145+
with open(assets_file_path, 'r') as file:
146+
lines = file.readlines()
147+
total_scenes_line = next(line for line in lines if "Total Scenes" in line)
148+
num_images = int(total_scenes_line.split(': ')[1].strip()) # Extract the number of scenes
149+
except Exception as e:
150+
logging.error(f"Failed to read or parse assets.txt: {str(e)}")
151+
exit(1)
152+
153+
# Now num_images is set dynamically based on the total scenes
154+
print(f"Number of images to generate: {num_images}")
155+
156+
157+
for i in range(num_images):
158+
# Randomly choose between the predefined prompts
159+
chosen_prompt = random.choice(prompts)
160+
print(f"Using original prompt: {chosen_prompt}") # Debugging line to check which prompt is chosen
161+
162+
# Improve the prompt using the API
163+
improve_payload = {"prompt": chosen_prompt}
164+
improve_response = requests.post(prompt_improve_url, json=improve_payload, headers=headers)
165+
166+
# Wait for 5 seconds to ensure the prompt has been improved
167+
time.sleep(5)
168+
169+
if improve_response.status_code == 200:
170+
response_json = improve_response.json()
171+
improved_prompt = response_json.get('promptGeneration', {}).get('prompt', None)
172+
173+
# Debugging line to check the full response from the API
174+
print("API Improve Response:", json.dumps(response_json, indent=4))
175+
176+
# Check if the prompt was actually improved
177+
if improved_prompt is None or improved_prompt == chosen_prompt:
178+
print("No improvement on the prompt. Stopping the process.")
179+
continue # Skip this iteration or use `break` to stop the entire loop
180+
181+
print(f"Improved prompt: {improved_prompt}") # Debugging line to check the improved prompt
182+
else:
183+
print("Failed to improve prompt, using base prompt.") # Debugging line for failure case
184+
print("HTTP Status:", improve_response.status_code, "Response:", improve_response.text) # Additional debugging information
185+
continue # Skip this iteration or use `break` to stop the entire loop
186+
187+
# Payload for the POST request to generate images
188+
payload = {
189+
"height": 1024,
190+
"prompt": improved_prompt, # Ensure this uses the improved prompt
191+
"modelId": "aa77f04e-3eec-4034-9c07-d0f619684628",
192+
"width": 576,
193+
"alchemy": True,
194+
"photoReal": True,
195+
"photoRealVersion": "v2",
196+
"presetStyle": "CINEMATIC",
197+
"num_images": 1
198+
}
199+
200+
# Make the POST request to generate images
201+
post_response = requests.post(post_url, headers=headers, json=payload)
202+
if post_response.status_code == 200:
203+
post_data = post_response.json()
204+
print("API Response:", json.dumps(post_data, indent=4))
205+
generation_id = post_data['sdGenerationJob']['generationId']
206+
max_attempts = 20
207+
attempts = 0
208+
while attempts < max_attempts:
209+
get_url = f"https://cloud.leonardo.ai/api/rest/v1/generations/{generation_id}"
210+
get_response = requests.get(get_url, headers=headers)
211+
if get_response.status_code == 200:
212+
get_data = get_response.json()
213+
if get_data['generations_by_pk']['status'] == 'COMPLETE':
214+
if get_data['generations_by_pk']['generated_images']:
215+
# Use the folder_path from earlier in the script
216+
if not os.path.exists(folder_path):
217+
os.makedirs(folder_path)
218+
for img in get_data['generations_by_pk']['generated_images']:
219+
image_url = img['url']
220+
image_response = requests.get(image_url)
221+
if image_response.status_code == 200:
222+
image_filename = image_url.split("/")[-1]
223+
with open(f'{folder_path}/{image_filename}', 'wb') as f:
224+
f.write(image_response.content)
225+
print(f"Image downloaded successfully: {image_url}")
226+
break
227+
else:
228+
print("No images generated yet, checking again...")
229+
elif get_data['generations_by_pk']['status'] == 'FAILED':
230+
print("Generation failed.")
231+
break
232+
else:
233+
print("Failed to retrieve generation details, HTTP Status:", get_response.status_code)
234+
time.sleep(15)
235+
attempts += 1
236+
if attempts == max_attempts:
237+
print("Maximum attempts reached, generation may still be processing.")
238+
else:
239+
print("Failed to generate images, HTTP Status:", post_response.status_code, "Response:", post_response.text)
240+
241+
# Step 11: Upload all images to S3 once they are all downloaded
242+
for local_image_path in local_image_paths:
243+
image_filename = os.path.basename(local_image_path)
244+
s3_key = f"{folder_name}/{image_filename}"
245+
s3_client.upload_file(local_image_path, BUCKET_NAME, s3_key)
246+
s3_image_url = f"https://{BUCKET_NAME}.s3.amazonaws.com/{s3_key}"
247+
print(f"Image uploaded to S3 at URL: {s3_image_url}")
248+
249+
# Optionally save the S3 URL to assets.txt
250+
with open(os.path.join(folder_path, 'assets.txt'), 'a') as file:
251+
file.write(f"S3 Image URL: {s3_image_url}\n")

0 commit comments

Comments
 (0)