This is a community-generated profile. If you work at Albs Labs, to claim it.
DiscoverStartupsAlbs LabsJobs at Albs Labs: Explore current Opportunities
Albs Labs company logo
1-10 Employees

Jobs at Albs Labs

Filter by
Show jobs on
Team
Location
Type
Engineering

Member of Technical Staff - Speech Language Models

NewPosted 1 month ago

Member of Technical Staff - Speech Language Models

  • Develop and train our speech language models end to end, from ASR and TTS to full speech-to-speech, both adapter-based on top of our LLM and as end-to-end full-duplex models.
  • Research and evaluate audio tokenizers and neural codecs that balance quality, bitrate and latency.
Engineering

Member of Technical Staff - Vision-Language Models

NewPosted 1 month ago

Member of Technical Staff - Vision-Language Models

  • Develop and train our vision-language models end to end, from vision encoders and training targets to data pipelines for rich, generalizing representations.
  • Build real-time visual understanding for machines: streaming video, grounding and spatial reasoning at low latency.
Engineering

Member of Technical Staff - LLM Pre-Training

NewPosted 1 month ago

Member of Technical Staff - LLM Pre-Training

  • Work across the pre-training stack, including data pipeline, model architecture, model optimization, training and evaluation.
  • Develop and evaluate novel ideas across architecture, optimizers, and training algorithms.
Engineering

Member of Technical Staff - Data Engineering

NewPosted 1 month ago

Member of Technical Staff - Data Engineering

  • Build the data platform behind our models: ingestion, storage and processing for pre-training, post-training and evaluation. For small on-device models, data quality is one of the biggest levers on model quality.
  • Build and scale pipelines for petabyte-scale text corpora: crawling, extraction, deduplication,...
Engineering

Member of Technical Staff - LLM Post-Training

NewPosted 1 month ago

Member of Technical Staff - LLM Post-Training

  • Work on post-training of our LLMs end to end: data curation, reward models, SFT, preference optimization, and RL methods such as GRPO.
  • Post-train models that run on-device: on-policy distillation from larger teachers, and alignment that holds up under quantization and tight latency budgets.