Job Overview:

Our team is seeking a highly motivated GAN/AI Engineering Intern who is passionate about leveraging cutting-edge Generative Adversarial Networks (GAN) techniques to build a model that is accurate in generating images based on text based prompts. The image generation has focussed styles, NLP inputs with which a line art based images are produced. This is state-of-art work which does not exist in the industry yet!

Responsibilities:

  • You will build a GAN model from scratch using PyTorch (or TensorFlow) with other framework & libraries from the ecosystem.
  • Research on key techniques, frameworks, mathematical models that would be necessary to build the model efficiently.
  • Identify or generate datasets which are illustrated images or SVG/Vector based data/images conforming to the requirements.
  • Work on preparing labeled dataset of illustrations and images/vectors producing quality dataset.
  • Work on understanding semantic context of words using NLP and translate into layers of Neural Network models.
  • Work on NLP based semantic understanding, text encoders/models and integrate them.
  • Train the model on Cloud GPUs and be cost aware.
  • Validate, Test and Measure the accuracy of the models.
  • Run multiple Epochs of the model by continually improving the model parameters, finetune and improve accuracy.
  • Produce an output of image based with highest probability of expected artistic images.
  • Required Skillset:

  • You are preferably a 2024 pass out with Data Science or related field as your major.
  • You have at least 1 internship of GAN based work in the past or have NLP based text encoding experience.
  • You are from IIT, NIT, VIT or any A+ institute in India - [This is an added advantage].
  • Solid understanding of GAN concepts, including image vectors, embeddings, Discriminator/Generator architecture, Text Encoding techniques.
  • Proficiency in Python programming.
  • PyTorch and Pyxxx frameworks.
  • Working with NLP libraries and frameworks such as SpaCy, or Hugging Face Transformers.
  • Ability to collaborate effectively within a team, while demonstrating strong, problem-solving skills.
  • Other requirements:

  • Work on an innovative & state-of-the-art project at the intersection of GAN/NLP.
  • Contribute to a product that addresses real-world challenges.
  • © 2023, Gettr Technologies (OPC) Pvt. Ltd. All Rights Reserved.