AI & ML interests

None defined yet.

Recent Activity

Organization Card

Qwen Forced Aligner - Polyglot Implementation

Multiple language implementations of the same forced alignment API for load balancing and educational purposes.

Available Languages

  • Python - Native PyTorch implementation
  • Node.js - JavaScript wrapper
  • Go - Go wrapper
  • More languages can be added as needed

Why Multiple Languages?

  1. Load Balancing - More spaces = more concurrent jobs. But its main purpose is for educational purposes.

  2. Learning Resource - See the same API in different languages. This helps developers understand how to implement Qwen in multiple languages.

  3. Flexibility - Choose your preferred language for contributing. All code is free to use under MIT license.

How It Works

All spaces expose the same HTTP API:

  • POST /align - Submit alignment job
  • GET /job/{id} - Check job status
  • GET /queue/status - Check if space is available

The ArchNemix controller automatically routes jobs to any available space.

Model

Uses Qwen-Forced-Alignment-0.6B for forced alignment.

Performance

  • ~5-15 seconds per job
  • 2 concurrent jobs per space
  • Total capacity increases with each language added

Part of ArchNemix

These spaces are part of the ArchNemix Text-to-Video pipeline.

Not intended for standalone use, but can still be used independently.

Contributing

Want to add another language implementation? Fork any existing space and translate it to your preferred language. Just maintain the same API contract.

License

MIT

models 0

None public yet

datasets 0

None public yet