Troubleshooting

This guide covers common problems you might encounter and how to fix them.

Setup & Installation Issues

Python version error: Requires Python 3.12

Symptom: Application crashes or shows syntax errors on startupCause: MoneyPrinter V2 requires Python 3.12 for compatibility with dependenciesSolution:

Check your Python version:
```
python3 --version
```

Install Python 3.12 if needed:

macOS (Homebrew):
```
brew install [email protected]
```

Ubuntu/Debian:

sudo add-apt-repository ppa:deadsnakes/ppa
sudo apt update
sudo apt install python3.12 python3.12-venv

Windows: Download from python.org

Recreate virtual environment:

rm -rf venv
python3.12 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

ModuleNotFoundError: No module named 'X'

Symptom: Import errors when running the applicationCause: Dependencies not installed or virtual environment not activatedSolution:

Activate virtual environment:

source venv/bin/activate  # Unix/macOS
.\venv\Scripts\activate   # Windows

Reinstall dependencies:
```
pip install -r requirements.txt
```
If using a specific module like faster-whisper:
```
pip install faster-whisper
```

Permission denied when running scripts

Symptom: bash: ./script.sh: Permission deniedSolution:Make scripts executable:

chmod +x scripts/*.sh
chmod +x scripts/*.py

Then run from project root:

bash scripts/setup_local.sh

Firefox Profile Issues

Firefox profile not found or invalid

Symptom: ValueError: Firefox profile path does not exist or is not a directoryCause: Incorrect path to Firefox profile in config.jsonSolution:Find your Firefox profile path:

macOS:
```
ls ~/Library/Application\ Support/Firefox/Profiles/
```
Look for a directory ending in .default-release Example: ~/Library/Application Support/Firefox/Profiles/abc123xyz.default-release
Linux:
```
ls ~/.mozilla/firefox/
```
Look for a directory ending in .default-release Example: ~/.mozilla/firefox/abc123xyz.default-release
Windows:
```
dir %APPDATA%\Mozilla\Firefox\Profiles\
```
Example: C:\Users\YourName\AppData\Roaming\Mozilla\Firefox\Profiles\abc123xyz.default-release

Update config.json:

{
  "firefox_profile": "/full/path/to/profile.default-release"
}

Verify it works:

python3 scripts/preflight_local.py

YouTube/Twitter login not persisting

Symptom: Browser asks to log in every timeCause: Using wrong Firefox profile or cookies not savedSolution:

Create a dedicated Firefox profile:
- Open Firefox
- Navigate to about:profiles
- Click “Create a New Profile”
- Name it (e.g., “MoneyPrinterV2”)
- Note the “Root Directory” path
Log into your accounts:
- Launch Firefox with this profile
- Log into YouTube, Twitter/X manually
- Save cookies/stay logged in
- Close Firefox
Update config.json with the new profile path
Test automation:
```
python3 src/main.py
```

Selenium WebDriver errors with Firefox

Symptom: WebDriverException: Message: Expected browser binary location, but unable to find binary in default locationCause: GeckoDriver or Firefox not foundSolution:

Ensure Firefox is installed:
```
firefox --version
```

Reinstall webdriver-manager:

pip install --upgrade webdriver-manager

Clear GeckoDriver cache:
```
rm -rf ~/.wdm
```
Run again - it will auto-download GeckoDriver

Ollama & LLM Issues

Ollama is not reachable

Symptom: [FAIL] Ollama is not reachable at http://127.0.0.1:11434Cause: Ollama service not runningSolution:

Check if Ollama is running:
```
curl http://127.0.0.1:11434/api/tags
```
Start Ollama:
- macOS: Ollama runs automatically if installed via Homebrew or app
- Linux:
  ollama serve
- Windows: Launch Ollama desktop app
Verify installation:
```
ollama --version
ollama list
```
Pull a model if none installed:
```
ollama pull llama3.2:3b
```

No models found on Ollama

Symptom: Warning about no available modelsSolution:

Pull recommended models:

ollama pull phi4:latest
ollama pull llama3.2:3b
ollama pull qwen3:14b

Verify models are installed:
```
ollama list
```
Re-run setup:
```
bash scripts/setup_local.sh
```
This will auto-detect and configure the best available model

Generated text is low quality or nonsensical

Symptom: LLM produces gibberish or off-topic contentCause: Model too small or poorly suited for taskSolution:

Use a larger, more capable model:

ollama pull qwen3:14b
# or
ollama pull deepseek-r1:32b

Update config.json:
```
{
  "ollama_model": "qwen3:14b"
}
```

Test generation:

ollama run qwen3:14b "Write a short video script about space exploration"

Ollama timeout or slow responses

Symptom: Long wait times or timeout errorsCause: Model too large for your hardware or heavy system loadSolution:

Use a smaller model:
```
ollama pull llama3.2:3b
```
Check system resources:
```
htop  # or Activity Monitor on macOS
```
Close unnecessary applications
Adjust Ollama settings (if running ollama serve):
```
OLLAMA_MAX_LOADED_MODELS=1 ollama serve
```

Image Generation Issues

Nano Banana 2 API errors (Gemini)

Symptom: [FAIL] nanobanana2_api_key is emptyCause: Missing or invalid Gemini API keySolution:

Get a Gemini API key:
- Visit Google AI Studio
- Create a new API key

Set in config.json:

{
  "nanobanana2_api_key": "AIzaSy..."
}

Or set as environment variable:

export GEMINI_API_KEY="AIzaSy..."

Verify connectivity:
```
python3 scripts/preflight_local.py
```

Image generation fails or returns no image

Symptom: [WARN] Nano Banana 2 did not return an image payloadCause: API quota exceeded, model unavailable, or malformed promptSolution:

Check API quota:
- Visit Google Cloud Console
- Ensure you haven’t exceeded free tier limits

Verify model availability: Update config.json to use stable model:

{
  "nanobanana2_model": "gemini-3.1-flash-image-preview"
}

Simplify prompts: Avoid overly complex or NSFW content that might be filtered

Video Generation Issues

ImageMagick not found - subtitle rendering fails

Symptom: [WARN] imagemagick_path is not set to a valid executable pathCause: ImageMagick not installed or path not configuredSolution:

Install ImageMagick: macOS:
```
brew install imagemagick
which magick  # Note this path
```
Ubuntu/Debian:
```
sudo apt install imagemagick
which convert  # Note this path
```
Windows:
- Download from imagemagick.org
- Install and note installation path (e.g., C:\Program Files\ImageMagick-7.1.0-Q16\magick.exe)

Update config.json:

{
  "imagemagick_path": "/usr/local/bin/magick"
}

Verify:
```
python3 scripts/preflight_local.py
```

MoviePy errors during video rendering

Symptom: OSError: MoviePy Error: creation of None failed because of the following errorCause: Missing codecs, ffmpeg issues, or insufficient disk spaceSolution:

Ensure ffmpeg is installed:

ffmpeg -version

If not installed:

# macOS
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

Check disk space:
```
df -h .
```
Videos are written to .mp/ directory
Lower thread count in config.json:
```
{
  "threads": 1
}
```
Try regenerating: Delete partial files in .mp/ and run again

Audio/video sync issues

Symptom: Subtitles or images don’t align with narrationCause: Timing calculation errors or TTS duration mismatchSolution:

Regenerate the video: Audio duration is calculated dynamically; retry generation
Check TTS configuration: Ensure tts_voice in config.json is valid:
```
{
  "tts_voice": "Jasper"
}
```
Manually adjust timing: Edit src/classes/YouTube.py:552 (combine method) to tweak duration calculations

Speech & Subtitle Issues

faster-whisper import error

Symptom: ModuleNotFoundError: No module named 'faster_whisper'Cause: faster-whisper not installed (required for local STT)Solution:

pip install faster-whisper

If you encounter installation issues:

Use AssemblyAI instead:

{
  "stt_provider": "third_party_assemblyai",
  "assembly_ai_api_key": "your-key-here"
}

Get AssemblyAI API key:
- Visit assemblyai.com
- Sign up for free tier

Subtitles not appearing in video

Symptom: Generated video has no subtitlesCause: ImageMagick not configured or STT failureSolution:

Check ImageMagick:
```
python3 scripts/preflight_local.py
```
Enable verbose logging:
```
{
  "verbose": true
}
```
Re-run and check for error messages
Verify STT provider: Ensure stt_provider is set correctly:
```
{
  "stt_provider": "local_whisper"
}
```
Check .mp/ directory: Look for generated .srt files - if missing, STT failed

Whisper model not found or slow

Symptom: Long wait times during subtitle generationCause: Whisper model downloading or too large for hardwareSolution:

Use a smaller model:

{
  "whisper_model": "tiny",
  "whisper_device": "cpu",
  "whisper_compute_type": "int8"
}

Available models: tiny, base, small, medium, large

Pre-download models: Models are auto-downloaded on first use but you can pre-cache:

from faster_whisper import WhisperModel
model = WhisperModel("base", device="cpu")

Upload & Automation Issues

YouTube upload fails at file picker

Symptom: Video doesn’t upload, browser stuck at upload screenCause: Element selectors changed or incorrect video pathSolution:

Check video file exists:
```
ls -lh .mp/*.mp4
```
Disable headless mode for debugging:
```
{
  "headless": false
}
```
Watch browser interactions to see where it fails
Update Selenium selectors: YouTube’s HTML changes frequently. Check src/classes/YouTube.py:703 (upload_video method) and update element selectors if needed
Ensure logged in: Open Firefox with your profile manually and verify YouTube login persists

Twitter/X post button not found

Symptom: RuntimeError: Could not find the Post button on X compose screenCause: X/Twitter UI changed or session expiredSolution:

Verify login: Open Firefox with profile and check if logged into X
Update selectors: Check src/classes/Twitter.py:73 (post method) for current selectors

Increase wait time: Edit src/classes/Twitter.py:71:

self.wait = WebDriverWait(self.browser, 60)  # Increase from 30

CRON jobs not running

Symptom: Scheduled automation doesn’t executeCause: Incorrect cron configuration or script path issuesSolution:

Test manual execution:

python3 src/cron.py youtube <account-id>

Check cron logs:

# macOS/Linux
tail -f /var/log/cron.log
# or
grep CRON /var/log/syslog

Use absolute paths in crontab:

0 */6 * * * cd /full/path/to/MoneyPrinterV2 && /full/path/to/venv/bin/python src/cron.py youtube account-id

Verify crontab syntax:
```
crontab -l
```

Performance Issues

Video generation is very slow

Symptom: Takes 10+ minutes to generate a short videoCause: Slow LLM, image generation, or video encodingSolution:

Use faster models:
- LLM: llama3.2:3b or phi4:latest
- Whisper: tiny or base
Reduce thread count (counterintuitively can help on some systems):
```
{
  "threads": 1
}
```
Limit script length:
```
{
  "script_sentence_length": 3
}
```
Profile bottlenecks: Enable verbose mode and time each step:
```
time python3 src/main.py
```

High memory usage or crashes

Symptom: System freezes or out-of-memory errorsCause: Large models loaded simultaneouslySolution:

Close other applications
Use smaller models:
- Ollama: 3B parameter models
- Whisper: tiny or base

Limit Ollama concurrent models:

OLLAMA_MAX_LOADED_MODELS=1 ollama serve

Process videos one at a time: Don’t run multiple automation jobs in parallel

Debugging Tips

Enable Verbose Logging

{
  "verbose": true
}

This prints detailed information about each step.

Check Log Files

MoneyPrinter V2 doesn’t create log files by default, but you can redirect output:

python3 src/main.py 2>&1 | tee moneyprinter.log

Inspect Generated Files

All temporary files are in .mp/ directory:

ls -lh .mp/

Check for:

.png - Generated images
.wav - TTS audio
.srt - Subtitle files
.mp4 - Final videos

Test Individual Components

# Test LLM
from src.llm_provider import generate_text
print(generate_text("Write a short poem about AI"))

# Test image generation
from src.classes.YouTube import YouTube
yt = YouTube(...)
yt.generate_image("A futuristic cityscape at sunset")

Use Python Debugger

# Add to any file
import pdb; pdb.set_trace()

Then step through code interactively.

Getting Help

If you’re still stuck:

Check existing issues: GitHub Issues
Search documentation: This site covers most scenarios
Ask on Discord: Fuji Community Discord
Open a new issue: Provide:
- Operating system and Python version
- Full error message and stack trace
- Relevant config.json (redact API keys)
- Steps to reproduce

Useful diagnostic info:

python3 --version
ollama --version
ffmpeg -version
magick --version
pip list | grep -E "moviepy|selenium|faster-whisper"

Include this output when asking for help.

Get Started

Features

Configuration

Advanced

Setup & Installation Issues

Firefox Profile Issues

Ollama & LLM Issues

Image Generation Issues

Video Generation Issues

Speech & Subtitle Issues

Upload & Automation Issues

Performance Issues

Debugging Tips

Enable Verbose Logging

Check Log Files

Inspect Generated Files

Test Individual Components

Use Python Debugger

Getting Help

Build docs developers (and LLMs) love

Get Started

Features

Configuration

Advanced

​Setup & Installation Issues

​Firefox Profile Issues

​Ollama & LLM Issues

​Image Generation Issues

​Video Generation Issues

​Speech & Subtitle Issues

​Upload & Automation Issues

​Performance Issues

​Debugging Tips

​Enable Verbose Logging

​Check Log Files

​Inspect Generated Files

​Test Individual Components

​Use Python Debugger

​Getting Help

Build docs developers (and LLMs) love

Setup & Installation Issues

Firefox Profile Issues

Ollama & LLM Issues

Image Generation Issues

Video Generation Issues

Speech & Subtitle Issues

Upload & Automation Issues

Performance Issues

Debugging Tips

Enable Verbose Logging

Check Log Files

Inspect Generated Files

Test Individual Components

Use Python Debugger

Getting Help