Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

OCC Live #5

Open
ArpitSureka opened this issue Oct 17, 2024 · 0 comments
Open

OCC Live #5

ArpitSureka opened this issue Oct 17, 2024 · 0 comments

Comments

@ArpitSureka
Copy link
Collaborator

ArpitSureka commented Oct 17, 2024


slug: occ-live
date: 04-Oct-2024
summary: summary
readingTime: 2min
image: https://www.tradersmagazine.com/wp-content/uploads/2024/02/Depositphotos_340767808_L-1024x683.jpg

Introduction to the interface

Hi developers! Welcome to OCC- your one stop discovery interface to find compute for projects in under 5 min. How?

Step 1:

Search by use case, provider or configuration needs

Step 2:

Click on the instance of your choice.

Step 3:

Redirect to your provider and spin up your instance.

Why OCC?

If you normally find your compute on the big 3 cloud providers you know how expensive it can be. It is often difficult to understand their complex instantiation process and having to pick every component of your instance takes up too much time. If you are a Kaggler or Colab user, customer support is difficult and storage and GPU are not seamless to upgrade.

OCC is made for you if any of this relates to you. Whether you are doing a passion project, completing your Deep Learning or Data Analytics projects, OCC is the place you can find your compute. We pride ourselves on being student friendly and hope to assist you on your journey here.

Without further ado, find and instantiate your compute, faster than you read this, Navigate to .
CoDMAV CodeJAM Hackathon Manual
CoDMAV hackathon powered by Open Cloud Compute
The Open Cloud Compute team is partnering with the Centre of Data Modelling Analytics and Visualisation (CoDMAV- PESU) to provide high quality compute to students and pilot OCC’s discoverability platform.
About the product
The platform allows students to discover Indian compute providers who are collaborators on the project. The compute is offered in the form of consumable services (Llama3 8B, Mistral 7B or ComfyUI hosted models) and notebooks (Jupyter, Pytorch and Tensorflow) with pre-installed libraries. All available provider compute offerings can be searched and filtered by the students. Considering the limited experience of many users, and large setup times of existing platforms that are too complicated, the platform abstracts away the details and provides a use-case based search feature. You can search, choose, instantiate and run any instance in under 5 minutes.

User journey: Divya, a college student, wants to build an online shopping platform for pet outfits. She wants to add a feature allowing a user’s pet to try on outfits. She recognises this needs image generation capabilities. Divya opens the OCC browser and chooses in her use case: image generation. The search identifies that ComfyUI is the best match tool and returns the search results to Divya who clicks and navigates to the provider instance, authenticates herself and spins up her instance.

The hackathon’s compute will come from OCC and students will be expected to use the product to build their projects.

Decorum

Problem statements

The problem statements can lie in the following tracks:

  1. Text Analytics and Natural Language Processing (NLP):
    Harness the power of Large Language Models (LLMs) like Llama 3: 8B and Mistral: 7B, or other open-source language models, to develop innovative solutions and products that push the boundaries of text processing and understanding.

Example problem statements:

  • Develop an AI-powered writing assistant that helps students improve their essay structure and content.
  • Create a multilingual chatbot for customer service that can understand and respond to queries in various languages.
  • Build a sentiment analysis tool for social media posts to help businesses understand public opinion about their products.
  1. Video, Image Analytics, and Computer Vision:
    Explore the cutting-edge capabilities of generative AI through frameworks like Stable Diffusion and ComfyUI, or train models from scratch to tackle complex visual computing challenges. Focus on image generation, segmentation, classification, or other innovative use cases.

Example problem statements:

  • Design an AI-powered fashion designer that generates unique clothing patterns based on user preferences.
  • Develop a real-time object detection system for autonomous vehicles using computer vision techniques.
  • Create an AI art restoration tool that can reconstruct damaged or faded portions of historical paintings.
  1. Big Data Analytics:
    Leverage up to 100GB of storage with accompanying Jupyter notebooks to build sophisticated analytics dashboards. Apply advanced data science techniques to extract meaningful insights and create value for users across various domains.

Example problem statements:

  • Develop a predictive maintenance system for industrial equipment using IoT sensor data and machine learning algorithms.
  • Create a personalised health recommendation engine that analyses user data to suggest lifestyle improvements.
  • Build a financial market trend analyzer that processes historical data to predict future market movements.
  1. Open Track:
    Unleash your creativity by combining any or all of the provided tools and technologies – LLMs, Stable Diffusion, PyTorch, TensorFlow, or Jupyter – to bring your most ambitious project ideas to life.

Example problem statements:

  • Develop an AI-powered virtual reality game that generates dynamic storylines and characters based on player interactions.
  • Create a multimodal AI assistant that can understand and respond to text, voice, and visual inputs for enhanced accessibility.
  • Build a comprehensive climate change impact analyzer that combines satellite imagery, weather data, and scientific reports to predict and visualise environmental changes.

We would also like to highlight outstanding projects built by the students on the website and will coordinate the same post the hackathon result announcements.

Agenda

Submission guidelines

Once you have built your project. Push your code to github as an open source repository. Submit the link here and give us some time to acknowledge your response.

## What happens if you win? Prize money: 1st place: Rs. 15,000/- 2nd place: Rs. 10,000/- 3rd place: Rs. 5,000/- For all: Goodies and the experience

SOS: Ask for help!

The CoDMAV and OCC teams are here to help you in any way. If you run into any, any issues fill out this form and we will find you!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant