Kaggle (@kaggle) 's Twitter Profile
Kaggle

@kaggle

The world's largest community of data scientists. Join us to compete, collaborate, learn, and share your work.

ID: 80422885

linkhttp://kaggle.com calendar_today06-10-2009 22:42:08

4,4K Tweet

288,288K Followers

277 Following

Kaggle (@kaggle) 's Twitter Profile Photo

🤖 Now on #KaggleModels! Part of the cohere and Cohere For AI family, Command R Plus is a multilingual (10 languages) model with advanced capabilities. This 104B parameter model includes RAG and tool automation and is optimized for reasoning, summarization, and question

Kaggle (@kaggle) 's Twitter Profile Photo

🤖 Now on #KaggleModels!   AI at Meta's largest openly available foundation model, Llama 3.1, excels in synthetic data generation and distillation with its 405B parameters, supporting 8 languages and a 128K context window. Also launching 7B and 80B models!   👉 Learn more:

Darek Kłeczek (@dk21) 's Twitter Profile Photo

Notes from LMSYS - Chatbot Arena Human Preference Predictions Kaggle competition top solutions (I didn’t participate). Goal is to predict which LLM response to a user prompt is better (as judged by Chat Arena users). Needs to be optimized to run for limited time on Kaggle (2xT4

Kaggle (@kaggle) 's Twitter Profile Photo

🤖 Now on #KaggleModels! Trained on thousands of hours of coral reef recordings, SurfPerch is a domain-adapted model for the classification of sounds in coral reefs. Use it as a standard TF 2.0 SavedModel. kaggle.com/models/google/…

Kaggle (@kaggle) 's Twitter Profile Photo

👇Check out this excellent distillation of insights from the lmsys.org competition by Darek Kłeczek. Participants built models to predict which responses users would prefer in a head-to-head LLM chatbot battle. Find the complete list of solution write-ups on the leaderboard here:

Kaggle (@kaggle) 's Twitter Profile Photo

📚 Check out this new starter notebook for ISIC 2024 Skin Cancer Detection competition. You'll learn how to train and deploy a deep learning model using EfficientNetV2 with KerasCV on skin lesion data. kaggle.com/code/awsaf49/i…

Kaggle (@kaggle) 's Twitter Profile Photo

📣 Competition Launch Alert! UM - Game-Playing Strength of MCTS Variants hosted by Maastricht University 🎯 Predict how MCTS variants perform against each other in different games. 💰 $50,000 Prize Pool ⏰ Entry Deadline: 11-25-24 Learn more 👇 kaggle.com/competitions/u…