ML Engineer- 6+Years Experience

Builds and operationalizes the LoRA fine-tuning pipeline that adapts the frozen Foundation base model. Responsible for implementing the parameter-efficient (LoRA) training loop, data preprocessing and tokenization, training-run execution and checkpointing, and packaging adapters for portable, repeatable deployment.

Key responsibilities
Implement the LoRA fine-tuning pipeline (adapter config, training loop, hyperparameter setup) against the frozen base model.
Build data preprocessing, tokenization, and train/validation splits for the clinical corpus.
Execute, monitor, and checkpoint training runs; track experiments and manage model/adapter versioning.
Package LoRA adapters as portable artifacts and integrate them into the inference/serving path.
Collaborate with AWS Data Scientists on evaluation harness and model-quality metrics.

Required skills:
Python, PyTorch, Hugging Face (Transformers/PEFT), LoRA/parameter-efficient fine-tuning, LLM training & evaluation, experiment tracking (e.g., MLflow/Weights & Biases).