Job description
Role Summary
Optimize and fine-tune large language models (GPT-4o, Claude, Llama-3, etc.) for client-specific tasks in English, Arabic, Urdu, and Bengali.
Key Responsibilities
Collect, clean, and annotate domain data sets.
Design fine-tuning and reinforcement-learning-from-human-feedback (RLHF) pipelines.
Benchmark model performance, latency, and cost.
Package and deploy models via Azure ML or Amazon Bedrock.
Must-Have Qualifications
3+ years in NLP or ML engineering.
Hands-on with Hugging Face Transformers, PEFT/LoRA, RLHF libraries.
Strong Python and PyTorch.
Experience with GPUs or distributed training (e.g., DeepSpeed).
Preferred
Prior work on Arabic or South-Asian language models.
MLOps exposure (Kubeflow, MLflow).
Engagement: Project-based, 2040 hrs/week, remote.
Job Type: Full-time
Work Location: On the road
Originally posted on Himalayas
Who can apply
Eligible countries: United States. Accepted UTC offsets: UTC-10, UTC-9, UTC-8, UTC-7, UTC-6, UTC-5, UTC+14. Review the full description for employer-specific work authorization, residency and schedule requirements.