No description available

A 3D web based interactive astronomical and biological database built for Nasa Space Apps Challenge 2025

In this repository, DeepSeek model architecture is created and trained on cosmopedia dataset

GRIFFURE.ORG's UNAI :: Almanax/Cosmopedia

No description available

Comprehensive EDA of Cosmopedia-100k dataset optimized for Mixture-of-Experts (MoE) language models' training efficiency. Features topic clustering, quality assessments. Includes interactive visualizations, and implementation guides for training efficiency improvemen

No description available

Cosmopedia: Gateway to the Universe Unified space data platform featuring rockets, astronauts, planets, agencies, telescopes, and AI chat - all accessible through modern web interface.

DeepSeek V3 replica trained on Cosmopedia v2 dataset

Architecture: Explains Recursive Blocks, SwiGLU, and MLA. Usage: How to run train.py (CPU vs GPU) and chat.py. Data: Credits the datasets (TinyGSM, Cosmopedia, etc.) and explains the "Clean Filter".

No description available

A wikipedia about cosmology

No description available

SMOLLM trained on cosmopedia

No description available

Django Rest Framework

Cosmopedia is a React Native app designed to provide detailed information about the planets, galaxies, black holes, and other celestial bodies in the universe. It serves as an educational resource for anyone interested in exploring the wonders of space, offering comprehensive and accessible content about the cosmos.

110M gpt model trained on cosmopedia "stories" subset with one H100 gpu

CosmoPedia is an interactive web application that lets users explore the universe through NASA's Astronomy Picture of the Day (APOD) API. Users can view stunning images and videos from space, along with detailed explanations of celestial phenomena, all within a visually immersive cosmic interface.

코스모피디아의 템플릿에 들어갈 정보를 입력하면 템플릿을 자동으로 작성합니다.

AR encyclopedia of all planets of the solar system, using vuforia ground plane detection and unity

Phi-4 was fine-tuned by cosmopedia dataset (https://huggingface.co/datasets/HuggingFaceTB/cosmopedia/viewer/auto_math_text/train?row=25)

SmolLM2-135M trained on Cosmopedia v2 for 5000 steps using PyTorch

SMOLLM2 Model trained from scratch using Shakespeare text, also added implementation to train on cosmopedia corpus

A replica of smollm-2 135M model made and trained from scratch. Dataset used is the Cosmopedia v2 within the smollm corpus dataset. Model trained till 10000 steps.

🧠 LLM Playground : Une exploration interactive des Grands Modèles de Langage (LLM). Comprend un modèle de 39.8M de paramètres entraîné from scratch (TinyGPT) et deux versions fine-tunées sur Wikipedia et des manuels scolaires (Cosmopedia).

Experimental GPT-2 scale (~124M param) LLM trained from scratch on Google Colab. Trained on C4, Cosmopedia/Alpaca/Python mix. Includes full training pipeline, mixed dataset loader with Colab-resilient checkpointing, and log analysis tools. Honest write-up of what went wrong.