Job Description
- Train state-of-the-art neural network models using publicly available datasets.
- Quantize the models using internal tools and optionally fine-tune the network architecture based on quantization results.
- Analyze models that do not quantize well, determine the reason(s) for the low quantization accuracy, and propose improvements.
Must-have:
- Candidates are expected to have recently graduated with an M.S. in Computer Science or a related field with specialization in Neural Networks.
- Strong experience training a multitude of neural network architectures using PyTorch/TensorFlow.
- Strong experience in curating datasets and augmenting datasets to meet accuracy requirements.
- Strong programming skills in C++ and Python.
- Strong background in algorithms and data structures.
- Good understanding of processor architectures.
Highly desirable:
- Strong understanding of machine learning algorithms such as matrix multiplication and convolution.
- Can-do attitude; Curious, creative, and good at solving problems; Execution and results oriented; Self-driven; Thinks Big and is highly accountable; Good communication skills.