Pruning-oriented training of Deep Neural networks
Master projects/internships - Leuven | More than two weeks ago
Develop new DNN training algorithms for more efficient AI.
IMEC is seeking talented individuals join its Compute System Architecture department in its endeavors to explore innovative training strategies for Deep Neural Networks. Participate in the efforts of our talented team of researchers to shape the future of AI computing.
Responsibilities
For this internship, you will be required to develop and evaluate a new, arithmetically guided pruning-based training method for Deep Neural Networks. Taking inspiration from works such as Frankle et al. "Lottery ticket Hypothesis" and Ramanujan et al. "What's Hidden in a Randomly Weighted Neural Network?" your work will give fundamental insight into the training dynamics of Deep Neural Networks.
Profile
- You are currently pursuing your Master of PhD degree.
- You are a confident python developer and are comfortable with the related machine learning libraries (i.e. Pytorch).
- You have an interest in computer architectures and are knowledgeable about digital arithmetic and number representations.
Type of project: Internship, Thesis
Required degree: Master of Science
Required background: Computer Science, Electrotechnics, Electrical Engineering
Duration: 6 months
Supervisor: Peter Vrancx
Supervising scientist(s): For more information on this topic, please contact Nathan Laubeuf (Nathan.Laubeuf@imec.be)
Imec allowance will be provided for students studying at a non-Belgian university.