Lamb Labs is seeking an exceptional ML engineer or researcher to tackle extreme low-bit language model inference. You’ll push 1–2‑bit quantization while preserving model quality, spanning quantization, activation handling, KV-cache optimization and training-related fine‑tuning.
You’ll collaborate with our hardware team to map models to custom silicon and evaluate novel representations for efficiency. You’ll work across PTQ, QAT, distillation and retraining where needed, aiming to make inference
Staff Engineer, Extreme Low-Bit LLM Inference in London employer: Lamb Labs
Lamb Labs provides a unique opportunity for professionals to work on groundbreaking AI technologies in a collaborative and research-focused environment. The company's commitment to innovation and efficiency in AI inference makes it an attractive employer for those looking to advance their careers in this exciting field.