Close Menu
    Trending
    • Mommies, Nannies, Au Pairs, and Me: The End Of Being A SAHD
    • Building Essential Leadership Skills in Franchising
    • History of Artificial Intelligence: Key Milestones That Shaped the Future | by amol pawar | softAai Blogs | Jun, 2025
    • FedEx Deploys Hellebrekers Robotic Sorting Arm in Germany
    • Call Klarna’s AI Hotline and Talk to an AI Clone of Its CEO
    • A First-Principles Guide to Multilingual Sentence Embeddings | by Tharunika L | Jun, 2025
    • Google, Spotify Down in a Massive Outage Affecting Thousands
    • Prediksi Kualitas Anggur dengan Random Forest — Panduan Lengkap dengan Python | by Gilang Andhika | Jun, 2025
    Finance StarGate
    • Home
    • Artificial Intelligence
    • AI Technology
    • Data Science
    • Machine Learning
    • Finance
    • Passive Income
    Finance StarGate
    Home»Machine Learning»DeepSeed-R1 Local API & Chatbot in Action | by Bill Huang | Feb, 2025
    Machine Learning

    DeepSeed-R1 Local API & Chatbot in Action | by Bill Huang | Feb, 2025

    FinanceStarGateBy FinanceStarGateFebruary 6, 2025No Comments3 Mins Read
    Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
    Share
    Facebook Twitter LinkedIn Pinterest Email


    Set up Ollama by downloading from their official web site in HERE for macOS M collection or Home windows, or use the command line for Linux:

    curl -fsSL  | sh

    This installs the mannequin runner with computerized GPU acceleration (Metallic for Apple M-series, CUDA for NVIDIA).

    Consider Ollama as a instrument that helps you obtain the DeepSeek-R1 mannequin and host it domestically in your machine, permitting different purposes to make use of this mannequin.

    Select a mannequin variant based mostly in your {hardware}. Yow will discover a listing of DeepSeek-R1 variations at https://ollama.com/library/deepseek-r1. These variations are smaller than the unique DeepSeek-R1 mannequin as a result of they’re distilled variations — consider it like a instructor (the unique mannequin) passing information to college students (smaller fashions).

    Whereas these compact variations retain a lot of the unique mannequin’s capabilities, their smaller measurement permits for decrease latency and decreased {hardware} necessities. Typically, bigger fashions are extra highly effective, however for native internet hosting, select one that matches your GPU’s capability.

    And thankfully, there’s a compact mannequin, DeepSeek-R1-Distill-Qwen-1.5B, that makes use of solely ~1GB VRAM and might even run on an M1 MacBook Air with 8GB reminiscence. We’ll use this mannequin as our instance, however be happy to discover different bigger fashions listed at https://ollama.com/library/deepseek-r1

    Pull and host the DeepSeek-R1 mannequin with Ollama with this commandline:

    ollama run deepseek-r1:1.5b # Balanced pace/high quality (~1.1GB VRAM)

    If you wish to expertise the DeepSeek mannequin as a ChatGPT-like chatbot, you possibly can make the most of the UI right here — it really works completely with the domestically hosted Ollama mannequin.

    Deploy Open WebUI for a user-friendly chatbot expertise with docker:

    docker run -d -p 3000:8080 -v ollama:/root/.ollama -v open-webui:/app/backend/knowledge 
    --add-host=host.docker.inner:host-gateway --name open-webui ghcr.io/open-webui/open-webui:major

    Go to http://localhost:3000, create an account, and choose deepseek-r1:1.5b from the mannequin dropdown.

    Utilizing a neighborhood LLM with the Chat UI affords a number of advantages:

    • Quick access to attempt totally different open-source fashions
    • Offline and native chatting capabilities, permitting you to make use of ChatGPT-like companies to spice up productiveness even with out web entry
    • Enhanced privateness, since Ollama and Open WebUI preserve all chat interactions native, stopping delicate info leaks

    You probably have already hosted DeepSeek-R1 domestically, as a developer, you possibly can simply combine with the mannequin since Ollama exposes an OpenAI-compatible API at http://localhost:11434/v1

    from openai import OpenAI
    from openai import OpenAI

    # Configure the shopper to make use of Ollama's native endpoint
    shopper = OpenAI(
    base_url="http://localhost:11434/v1", # Ollama server tackle
    api_key="no-api-key-needed", # Ollama would not require an API key
    )

    from langchain_ollama import ChatOllama
    # Configure the Ollama shopper to make use of native endpoint
    llm = ChatOllama(
    base_url="http://localhost:11434", # Ollama server tackle
    mannequin="deepseek-r1:1.5b", # Specify the mannequin you host in Ollama
    )

    All associated code and setup may be discover right here:

    Right here’s a fast overview of the method:

    With Ollama and DeepSeek-R1, now you can run highly effective AI domestically with GPU acceleration, entry a ChatGPT-like interface by Open WebUI, and combine AI capabilities into your purposes utilizing commonplace APIs — all whereas sustaining privateness and dealing offline.



    Source link

    Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
    Previous ArticleSocial Links Launches Darkside AI Initiative to Address Cybercrime and Misinformation
    Next Article driving the next evolution of AI for business
    FinanceStarGate

    Related Posts

    Machine Learning

    History of Artificial Intelligence: Key Milestones That Shaped the Future | by amol pawar | softAai Blogs | Jun, 2025

    June 13, 2025
    Machine Learning

    A First-Principles Guide to Multilingual Sentence Embeddings | by Tharunika L | Jun, 2025

    June 13, 2025
    Machine Learning

    Prediksi Kualitas Anggur dengan Random Forest — Panduan Lengkap dengan Python | by Gilang Andhika | Jun, 2025

    June 13, 2025
    Add A Comment

    Comments are closed.

    Top Posts

    A Beginner’s Guide to Reinforcement Learning with PyTorch! | by Emrullah AYDOGAN | Apr, 2025

    April 3, 2025

    Rice Univ. Prof. Lydia Kavraki Elected to National Academy of Engineering for Research in Biomedical Robotics

    February 14, 2025

    Traveling Professionals: Add This MacBook Air to Your Carry-on for Less Than $200

    March 29, 2025

    Survey: 84% Say AI Won’t Replace Low-Code and No-Code Tools

    March 28, 2025

    These documents are influencing the DOGE-sphere’s agenda

    February 7, 2025
    Categories
    • AI Technology
    • Artificial Intelligence
    • Data Science
    • Finance
    • Machine Learning
    • Passive Income
    Most Popular

    5 Digital Marketing Statistics to Improve Your Law Firm’s Strategy in 2025

    February 5, 2025

    Top Python Libraries for Machine Learning | by Expert App Devs | Apr, 2025

    April 14, 2025

    Amplifying Creativity: Building an AI-Powered Content Creation Assistant — Part 3 | by Markell Richards | Apr, 2025

    April 7, 2025
    Our Picks

    Model Context Protocol (MCP): The Force Awakens | by Gourav Didwania | Mar, 2025

    March 28, 2025

    Myths vs. Data: Does an Apple a Day Keep the Doctor Away?

    February 6, 2025

    How to Get Promoted as a Data Scientist

    February 4, 2025
    Categories
    • AI Technology
    • Artificial Intelligence
    • Data Science
    • Finance
    • Machine Learning
    • Passive Income
    • Privacy Policy
    • Disclaimer
    • Terms and Conditions
    • About us
    • Contact us
    Copyright © 2025 Financestargate.com All Rights Reserved.

    Type above and press Enter to search. Press Esc to cancel.