2 months ago
Barcelona, SpainMid Level / Senior
H1B Sponsor
Responsibilities
- Architect, implement, pre-train, fine-tune, and post-train large-scale speech models.
- Independently lead small research projects and collaborate on larger initiatives.
- Design, run, and analyze scientific experiments to improve model understanding.
- Develop and enhance development tooling for team productivity.
- Contribute to the entire stack from low-level optimizations to high-level model design.
- Define data requirements and collaborate on data acquisition and quality.
- Design automated evaluations and conduct robustness and bias checks.
- Harden the training, evaluation, and inference pipeline to meet production SLAs.
- Partner with infrastructure for distributed training and productionize models.
- Contribute to safety and consent guardrails for responsible speech technology.
Requirements
- Exceptional experience with large scale audio models (>3B models and >500k hours data).
- Strong understanding and hands-on experience with transformer architectures and/or diffusion models.
- Experience with multi-node and multi-GPU distributed model training.
- Proven software engineering skills in building complex systems.
- Proficiency in PyTorch and performance optimization techniques.
- Experience shipping large scale speech/audio models to production.
- Background in working with large-scale ML data.
- Ability to iterate on data quality using subjective and objective signals.
- Notable publications or open-source contributions in speech/audio/ML.
- Experience with voice-cloning, speech-control, and voice-generation.