A general demo of the American Data Science Platform

Overview

American Data Science

American Data Science helps AI/ML Engineers and Data Scientists accelerate their notebook development workflow with pre-built GPU environments, compute, and AI agents.

Introduction

Getting started with powerful notebook servers on CPUs or H100s with a click of the button.

Quickstart

Launch your notebook servers on GPUs with cloud providers like Lambda Labs and DataCrunch at no additional cost.

Compute

A powerful AI assistant that generate and run code directly in your notebook servers

Agent Alph

Easily share your applications and services straight from your notebook servers.

Port Forwarding

A demo of extracting knowledge graphs from GPT-4o and Claude 3

LLM-based Knowledge Graph Generation

A demo of simulating chemical catalysts with GPUs

Meta's AI Research Chemistry

Use your API Key for authentication when using our API and SDKs.

Authentication

Get Environments

Get Compute

Get Servers

Create Server

Stop Server

Start Server

Delete Server

Get File

Upload File

Run Code

gpt-4o

gpt-4

gpt-3

claude3-5

claude3

gemini

mistral

perplexity

groq

fireworks

Open Source

Forum

API Reference

Dashboard

Support

Our environments are GPU-enabled and support PyTorch, Jax, RAPIDS, Transformers, and much more.

The AI environments are each tailored for specific AI tools and SDKs. They include tools for general AI development like PyTorch, as well as those specifically designed for working with AI models and tools from Anthropic, Cohere, Fireworks, Langchain, Mistral AI, OpenAI, Perplexity, and Replicate. Each environment provides the necessary clients, tools, and libraries for efficient development and interaction with the respective AI models and APIs.

The Computer Vision environments are designed for various computer vision tools and libraries. These include environments for general computer vision tasks, PyTorch Image Models (TIMM), OpenCV, and Pillow. Each environment is equipped with the necessary tools for efficient image processing, model training, and real-time computer vision tasks.

Computer Vision

The Data Science and Machine Learning environments are designed for a variety of data science tools and machine learning libraries. These include environments for xgboost, lightgbm, scikit-learn, scipy, pandas, numpy, matplotlib, seaborn, numba, and cupy. Each environment is equipped with the necessary tools for efficient data analysis, machine learning model training, and numerical computations.

Data Science and Machine Learning

The Data Visualization environments are designed for various data visualization tools and libraries. These include environments for matplotlib, seaborn, plotly, hvplot, datashader, dash, and more. Each environment is equipped with the necessary tools for efficient data visualization tasks, providing a streamlined experience for creating interactive, publication-quality graphs and web analytic applications.

Data Visualization

The Deep Learning and Tensors environments are designed for specific deep learning libraries and tools. These include environments for PyTorch, JAX, TensorFlow, Hugging Face, and Fast.ai, each equipped with the necessary tools for efficient deep learning and tensor computations. They provide a streamlined experience for tasks such as training neural network models, high-performance numerical computing, and large-scale machine learning.

Deep Learning and Tensors

The Model Deployment environments are designed for various model deployment tools and libraries. These include environments for Gradio, Modelbit, and Streamlit, each equipped with the necessary tools for efficient model deployment and user interface creation. They provide a streamlined experience for tasks such as creating customizable UI components around your models, working with the Modelbit package and ecosystem, and using Streamlit, an open-source app framework for Machine Learning and Data Science teams.

Model Deployment

The Vector Databases environments are tailored for specific vector database tools and clients, including vectordb, pinecone, milvus, weaviate, raft, faiss, chroma, and qdrant. Each environment is equipped with the necessary libraries and tools for efficient interaction with the respective vector databases. They facilitate tasks such as vector search, analytics, and clustering of dense vectors.

Vector Databases

The Natural Language Processing environments are designed for various NLP tools and libraries. These include environments for spacy, nltk, gensim, and flair, each equipped with the necessary tools for efficient natural language processing tasks. They provide a streamlined experience for tasks such as text analysis, language understanding, and topic modeling.

Get Started

Guides

Demos

Overview