ML Gradient Wars 2026

About

ML Gradient Wars 2026 is a 48-hour, paid-entry machine-learning battle. One round, one dataset, one leaderboard. The problem statement stays under wraps until 13 Nov, 10:00 AM IST, and the round closes on 15 Nov at 10:00 AM IST.

₹5 lakh in prizes, and every member of the top 3 teams is referred to The Data Foundation for a 3-month internship.

How it works

1. Register as a team of 1 to 4 members and pay the ₹1,999 entry fee per team, by 10 Nov, 11:59 PM IST.
2. Problem statement and dataset go live on 13 Nov, 10:00 AM IST in your team workspace under Resources, with a sample submission and a data dictionary.
3. The 48-hour live round: 13 Nov, 10:00 AM to 15 Nov, 10:00 AM IST. Upload predictions from the Leaderboard tab, up to 10 times a day. The leaderboard updates every 15 minutes.
4. Submit your code (GitHub repository or notebook link) from the Submission tab before the round closes.
5. Final ranking and code verification on 15 Nov, on the private part of the test set. Results at 5:00 PM IST.

Prizes

• 1st place: ₹2,00,000
• 2nd place: ₹1,50,000
• 3rd place: ₹1,00,000
• 4th place: ₹30,000
• 5th place: ₹20,000
• Top 3 teams: every team member is referred to The Data Foundation for a 3-month internship
• Certificates for every team that uploads a scored submission

Scoring

• The evaluation metric is announced with the problem statement.
• The live leaderboard updates every 15 minutes and is scored on a public part of the test set.
• The final ranking uses the hidden private part, based on each team's best public upload. Models that generalise win; overfitting the leaderboard does not.

Who can participate

• College students from any year or branch, and working professionals
• Teams of 1 to 4 members
• ₹1,999 entry fee per team

Prizes

• 1st place — ₹2L
• 2nd place — ₹1.5L
• 3rd place — ₹1L
• 4th place — ₹30K
• 5th place — ₹20K
• Top 3 teams: Every team member referred to The Data Foundation for a 3-month internship

Perks

• Every member of the top 3 teams referred to The Data Foundation for a 3-month internship
• Certificate of achievement for the top 3
• Certificate for every team that uploads a scored submission

Timeline

1. Registration and team formation (24 Oct 2026 – 10 Nov 2026): Register as a team of 1 to 4 members and pay the ₹1,999 entry fee per team to confirm your place. Registration closes on 10 Nov, 11:59 PM IST.
2. Problem statement and dataset release (13 Nov 2026 – 13 Nov 2026): The problem statement, training data, test file (without labels), a sample submission and the data dictionary go live in your team workspace under Resources at 10:00 AM IST on 13 Nov.
3. ML Gradient Wars: 48-hour live round (13 Nov 2026 – 15 Nov 2026): One round, one dataset, one leaderboard: 13 Nov, 10:00 AM to 15 Nov, 10:00 AM IST. Upload predictions from the Leaderboard tab, up to 10 times a day. The leaderboard updates every 15 minutes. Submit your code from the Submission tab before the round closes.
4. Final ranking and code verification (15 Nov 2026 – 15 Nov 2026): The leaderboard switches to the private part of the test set. The top teams' code is checked to make sure it reproduces their best upload.
5. Results (15 Nov 2026 – 15 Nov 2026): Winners announced at 5:00 PM IST on the hackathon page and by email.

Rules

1. Take part in a team of 1 to 4 members. Each person can be in only one team, with one account.
2. Each team pays a ₹1,999 entry fee (per team, not per person) when registering. A team is registered only once the fee is paid.
3. There is one 48-hour round: 13 Nov, 10:00 AM to 15 Nov, 10:00 AM IST. The problem statement and dataset are released at the start.
4. Train only on the dataset provided. Open-source libraries and publicly available pretrained weights are allowed; outside labelled data for the same task and hand-labelling the test set are not.
5. Upload predictions in the format given with the problem statement, up to 10 times per team per day (the limit resets at midnight IST). Files that fail validation do not use up an upload.
6. The live leaderboard uses a public part of the test set and updates every 15 minutes. The final ranking uses the hidden private part, based on each team's best public upload.
7. Before the round closes, submit your code (a GitHub repository or notebook link) from the Submission tab. Prize-winning teams must be able to reproduce their best upload from that code.
8. Sharing code, predictions or labels between teams is not allowed and leads to disqualification.
9. Internship referrals go to every member of the top 3 teams in the final ranking. The internship itself is offered at The Data Foundation’s discretion after their interview.
10. The organizers' decisions on verification and results are final.

ML Gradient Wars 2026: quick answers

What is ML Gradient Wars 2026?

ML Gradient Wars 2026 is an online ML competition with 3-month internship, organized by Hack4Hack. Top 3 teams: every member is referred to The Data Foundation for a 3-month internship, on top of the cash prizes. One 48-hour round on a live leaderboard.

When is ML Gradient Wars 2026?

ML Gradient Wars 2026 runs from 13 Nov 2026 to 15 Nov 2026, online. Registration is open until 10 Nov 2026. Stages: Registration and team formation (24 Oct 2026); Problem statement and dataset release (13 Nov 2026); ML Gradient Wars: 48-hour live round (13 Nov 2026); Final ranking and code verification (15 Nov 2026); Results (15 Nov 2026).

What is the last date to register for ML Gradient Wars 2026?

Registration for ML Gradient Wars 2026 closes on 10 Nov 2026 (11:59 pm IST). Register on Hack4Hack before then; teammates can join with your team code.

What is the prize money at ML Gradient Wars 2026?

ML Gradient Wars 2026 has ₹5,00,000 (₹5 lakh) in prizes. 1st place: ₹2,00,000 (₹2 lakh), 2nd place: ₹1,50,000 (₹1.5 lakh), 3rd place: ₹1,00,000 (₹1 lakh), 4th place: ₹30,000, 5th place: ₹20,000. ₹5 lakh cash + internship referrals at The Data Foundation.

Is ML Gradient Wars 2026 free?

No. ML Gradient Wars 2026 has a registration fee of ₹1,999, paid online after registering.

Who can participate in ML Gradient Wars 2026?

College students and working professionals can take part, in teams of 1 to 4. It is fully online, so you can join from anywhere.

Can I join ML Gradient Wars 2026 from outside India?

Yes. ML Gradient Wars 2026 runs fully online, so you can take part from any country unless its rules say otherwise. It starts 13 Nov, 10:00 am IST (13 Nov, 4:30 am UTC). The ₹1,999 entry fee is paid online in Indian rupees.

How do I get an internship through ML Gradient Wars?

Finish in the top 3. Every member of the top 3 teams is referred to The Data Foundation for a 3-month internship, on top of the cash prize.

How does ML Gradient Wars 2026 work?

One 48-hour online round, one dataset and one live leaderboard. The problem statement and dataset are released when the round starts; teams upload predictions up to 10 times a day and submit their code at the end for verification.

What is the team size for ML Gradient Wars?

Teams of 1 to 4. You can compete solo or with up to three teammates.

Who should join ML Gradient Wars?

Students and professionals who want to test their machine learning skills on a real dataset, win prize money and earn an internship referral, from anywhere, online.

Frequently asked questions

Is there a fee?

Yes. ₹1,999 per team (not per person), paid when you register, by UPI, card or netbanking. Your team is registered once the fee is paid.

Can I take part alone?

Yes. Teams can have 1 to 4 members, so you can compete solo.

How many rounds are there?

Just one: a 48-hour live round, 13 Nov, 10:00 AM to 15 Nov, 10:00 AM IST. The problem statement and dataset are released at the start.

What are the prizes?

A ₹5,00,000 prize pool: ₹2,00,000 for 1st, ₹1,50,000 for 2nd, ₹1,00,000 for 3rd, ₹30,000 for 4th and ₹20,000 for 5th. On top of that, every member of the top 3 teams is referred to The Data Foundation for a 3-month internship.

Is the internship guaranteed?

The referral is guaranteed for every member of the top 3 teams. The internship offer is decided by The Data Foundation after the interview.

How is it scored?

The evaluation metric is announced with the problem statement. The live leaderboard uses a public part of the test set; the final ranking uses the hidden private part.

How many times can we upload?

Up to 10 times per team per day. The limit resets at midnight IST, and uploads that fail validation do not count.

Do I need a GPU?

No. Free tiers such as Google Colab or Kaggle notebooks are enough.