Event Schedule
08.30 - 09.00: Registration
09.00 - 10.00: Reply Presentation
10.00 - 13.00: Hacking
13.00 - 14.00: Lunch
14.00 - 18.00: Hacking
18.00 - 18.45: Evaluation
18.45 - 19.00: Awards Presentation
About the Challenge
The Challenge for Hack Cambridge Spyder Lite has been divided into two main areas of Machine Learning (Computer Vision and Natural Language Processing), based on varying levels of difficulties. We want to give you as much creative freedom as possible!
Remember to submit your work once hacking ends, to compete for the prizes and receive feedback on your solutions from the experts of our sponsor company.
Get Started
- Find a dataset of images or text (depending on the chosen area of interest) and identify the data which you would like to generate.
- Links to high-quality data will be provided for images. For text, Kaggle.com is an ideal resource to use.
- As the field of NLP is diverse, using publicly available textual datasets can prove be useful when searching for data against a given use-case.
- Using Google Colab, explore your data and understand exactly what it is which you would like to generate.
- Take some time to prepare your data to feed into your pre-trained model
Useful resources and example notebooks can be found at this Google Drive link.
https://drive.google.com/drive/u/1/folders/1R4boDg9kBaxKKLLbnuGQKszeGuKgPSs1
Requirements
What to Build
Build an end-to-end solution which is able to generate the data in the chosen area.
The solution should effectively be a pipeline which is able to:
- Ingest the data.
- Pre-process the data.
- Feed the data into a Large Language Model.
- Output a generated image or texts.
What to Submit
Submit a zip file, containing all of the relevant materials used to complete the challenge.
The zip file must contain (at a minimum):
- A python notebook (.ipynb file) used for the attempt of the challenge
- A PowerPoint presentation which explains the steps, choices and results (a minimum of 3 slides are expected for adequate evaluation)
Prizes
£756 in prizes
3rd Generation Apple AirPods
(4)
We’re excited to announce that the first place prize for our hackathon is a set of 3rd generation Apple AirPods for each member of the winning team
This prize is kindly provided by our sponsor Reply.
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Fabio Ardossi
Reply Expert - Data Reply
Saif Saleem
Reply Expert - Data Reply
Alex Dmitrewski
Reply Expert - Data Reply
Charles Young
Reply Expert - Solidsoft Reply
James Smith
Reply Expert - Sprint Reply
Judging Criteria
-
TECHNICAL SOLUTION QUALITY (30)
The technical quality of the proposed solution and data generated (processes which the solution follows, how it could be further extended, realism, diversity and representativeness of the underlying distribution of features / classes in the dataset) -
BUSINESS FEASIBILITY (10)
How realistic is this idea in a business context? Develop your idea around a brand. How well has it been branded? -
NOVELTY AND CREATIVITY (30)
Originality of solution used to generate data, Creativity in approach taken (Number of models used, How intuitive input prompts are, Techniques used), Performance against test set, Understandability of results (Perplexity, Accuracy, etc.) -
TEAMWORK AND CLARITY OF PRESENTATION (15)
Clarity and Effectiveness of the team's presentation of their work (Ability to communicate the technical details and significance of their approach, Quality of any visual aids or demonstrations used) -
VISION FOR FUTURE EXTENSION (15)
What can be done to improve this solution? How can the project be improved?
Questions? Email the hackathon manager
Tell your friends
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.