Sign inSign up

xavierlam/subaligner

By xavierlam

•Updated 8 months ago

Synchronize and align subtitle timing to match a reference subtitle

Image
0

1.3K

xavierlam/subaligner repository overview

⁠Audio2Sub Docker Images

Audio2Sub is an all-in-one subtitle toolkit. Official Docker images provide easy deployment for all three tools: audio2sub (speech-to-text transcription), subtranslator (AI subtitle translation), and subaligner (AI subtitle timing synchronization).

⁠Quick Start

⁠audio2sub — Transcribe Audio to Subtitles

Significantly faster transcription with GPU:

# Using faster-whisper (default, recommended)
docker run --rm --gpus all \
  -v "$(pwd):/media" \
  xavierlam/audio2sub \
  video.mp4 -o video.srt --lang en

# Using OpenAI Whisper
docker run --rm --gpus all \
  -v "$(pwd):/media" \
  xavierlam/audio2sub:whisper \
  video.mp4 -o video.srt --lang en

# Using faster-whisper without GPU
docker run --rm -v "$(pwd):/media" xavierlam/audio2sub \
  video.mp4 -o video.srt --lang en

# Using Gemini API (no GPU needed)
docker run --rm -v "$(pwd):/media" \
  -e GEMINI_API_KEY=your_api_key_here \
  xavierlam/audio2sub:gemini \
  video.mp4 -o video.srt --lang en
⁠subtranslator — Translate Subtitles
# Translate Chinese subtitles to English (Gemini, default)
docker run --rm -v "$(pwd):/media" \
  -e GEMINI_API_KEY=your_api_key_here \
  xavierlam/subtranslator \
  video_zh.srt -s zh -d en -o video_en.srt

# Use Grok backend
docker run --rm -v "$(pwd):/media" \
  -e GROK_API_KEY=your_api_key_here \
  xavierlam/subtranslator \
  video_zh.srt -s zh -d en -o video_en.srt -t grok

# Use OpenAI backend
docker run --rm -v "$(pwd):/media" \
  -e OPENAI_API_KEY=your_api_key_here \
  xavierlam/subtranslator \
  video_zh.srt -s zh -d en -o video_en.srt -t openai
⁠subaligner — Align Subtitle Timing
# Align Chinese subtitles using English reference timing
docker run --rm -v "$(pwd):/media" \
  -e GEMINI_API_KEY=your_api_key_here \
  xavierlam/subaligner \
  -i chinese.srt -r english_reference.srt -o output.srt --src-lang zh --ref-lang en

# Use Grok backend
docker run --rm -v "$(pwd):/media" \
  -e GROK_API_KEY=your_api_key_here \
  xavierlam/subaligner \
  -i chinese.srt -r english_reference.srt -o output.srt -a grok

⁠Available Images

All images are available on Docker Hub.

⁠audio2sub
TagBackendDescription
xavierlam/audio2sub:latestfaster-whisperFast CTranslate2-based Whisper (Recommended)
xavierlam/audio2sub:faster-whisperfaster-whisperSame as latest
xavierlam/audio2sub:whisperwhisperOriginal OpenAI Whisper
xavierlam/audio2sub:geminigeminiGoogle Gemini API
⁠subtranslator
TagDescription
xavierlam/subtranslator:latestAI subtitle translator with Gemini, Grok, and OpenAI support
⁠subaligner
TagDescription
xavierlam/subaligner:latestAI subtitle aligner with Gemini, Grok, and OpenAI support

⁠Troubleshooting

⁠GPU Not Working

If GPU acceleration isn't working, verify your setup:

  1. Check NVIDIA drivers: Run nvidia-smi on your host machine. You should see your GPU listed.
  2. Verify Docker GPU access: Run docker run --rm --gpus all nvidia/cuda:11.8.0-base-ubuntu22.04 nvidia-smi. This should show the same GPU information inside Docker.
  3. Install NVIDIA Container Toolkit: If step 2 fails, you need to install the NVIDIA Container Toolkit⁠.
  4. Check Docker version: Ensure you have Docker 19.03+ which supports the --gpus flag.

⁠Resources

Tag summary

Content type

Image

Digest

sha256:40925c70b…

Size

60.6 MB

Last updated

8 months ago

docker pull xavierlam/subaligner