Deep Learning
November 13 – December 19, 2021
Kyiv, offline
Free school for junior and mid-level
ML/DL researchers & engineers
Application closed
Creator School
Application closed
AI HOUSE’s Deep Learning Creator School
is here to give knowledge and tools to pursue ML-powered content creation and transformation. It is a series of lectures and workshops aimed at junior and mid-level ML/DL researchers and engineers.
is here to give knowledge and tools to pursue ML-powered content creation and transformation. It is a series of lectures and workshops aimed at junior and mid-level ML/DL researchers and engineers.
6 weekends
offline
15 students
free course
You will learn
how to:
how to:
Select ideas
Validate Hypotheses
Research DL innovations
Train models
Deploy and scale solutions
Program:
deep learning content creation simply explained
6 weeks, Sat–Sun, 10 am–7 pm
Week 01
• Opening
• Analyzing your Kaggle competition results and finding ways to improve them
• Face swap for images: problem definition, theory, baseline solution
• Analyzing your Kaggle competition results and finding ways to improve them
• Face swap for images: problem definition, theory, baseline solution
Week 02
• Face swap for videos: problem-solving, evaluation metrics, consistency
• Neural networks optimization: during the training (Apex) and inference (JIT, Glow)
• Neural networks optimization: during the training (Apex) and inference (JIT, Glow)
Week 03
• Emotion swap: problem definition, dataset collection, baseline solution
• High-performance Python engineering (multiprocessing, GIL, shared memory, etc.)
• Course project start: ideas selection and discussion (in teams)
• High-performance Python engineering (multiprocessing, GIL, shared memory, etc.)
• Course project start: ideas selection and discussion (in teams)
Week 05
• Model deployment to production: architecture, ways to optimize cost and latency
• Model serving frameworks (Bentoml, TorchServe, Vertex.ai)
• Course project: moving to real products
• Model serving frameworks (Bentoml, TorchServe, Vertex.ai)
• Course project: moving to real products
Week 06
• Course project finalization, deploy to the cloud
• Course project results presentation
• Graduation
• Course project results presentation
• Graduation
Week 04
• Emotion swap: problem-solving, ways to improve, architecture limitations
• Product mindset essentials
• Course project: ideas validation, the first iteration results
• Product mindset essentials
• Course project: ideas validation, the first iteration results
(01)
we offer free education for those
who successfully pass
the selection process
who successfully pass
the selection process
Quiz on basic knowledge of machine learning
Admission:
Interview
(03)
Step
Technical task in a format of a Kaggle open competition
(02)
Step
take the quiz now.
take the quiz now.
take the quiz now.
take the quiz now.
Find your path
in deep learning
content creation
in deep learning
content creation
Application closed
take the quiz now.
Have questions?
Look no further. Our FAQ has got you covered
Who should take this course?
We encourage you to apply only if you have relevant hard skills and experience. The best candidates would be junior and mid-level machine learning and deep learning researchers and engineers who have at least 1 year of working experience. If you are a backend developer or a frontend developer who wants to switch to machine learning, you are also welcome to enroll.
What’s your tuition?
The Deep Learning Creator School is a free offline course. We don’t charge any fees.
What do I get once I graduate?
- Knowledge of deep learning content creation, transformation, and production
- Course certificate
- Selected students will receive job offers
What is the course schedule?
Starting November 13, we will be meeting onsite every Sat and Sun until December 19. The classes run at 10 am through 7 pm.
What will the education process be like?
The learning process will be 30% theory and 70% practical activities in teams. Our lecturers and mentors will guide you. Each weekend will be dedicated to a different topic (you can look up all the program topics above). Ideally, you will have a working prototype by the end of each Sunday so that no extra work will be required on weekdays.
Will the lectures be recorded?
No, we insist on an offline format of the course with no video recording.
What are the technical requirements and what software will we use?
- Hard skills: Python, Pytorch
- Software: Ubuntu, FFmpeg, Google Cloud ML Engine
- Hardware: laptop or PC with a GPU
How will we communicate with the lecturers?
We will create a group chat on Slack for the students and lecturers to communicate throughout the entire duration of the course.
If you would like more information about a specific topic, please email to
