Install Navarasa Model Locally - LLM for 15 Indian Languages

แชร์
ฝัง
  • เผยแพร่เมื่อ 7 ต.ค. 2024
  • This video shows how to locally install Navarasa 2.0, a Gemma 7B/2B instruction-tuned model in 15 Indian Languages.
    ▶ Become a Patron 🔥 - / fahdmirza
    #navarasa #navarasa2
    PLEASE FOLLOW ME:
    ▶ LinkedIn: / fahdmirza
    ▶ TH-cam: / @fahdmirza
    ▶ Blog: www.fahdmirza.com
    RELATED VIDEOS:
    ▶ How To Use Custom Dataset with Mixtral 8x7B Locally • How To Use Custom Data...
    ▶ Introduction to AWS Bedrock • Amazon Bedrock Introdu...
    ▶ Model huggingface.co...
    All rights reserved © 2021 Fahd Mirza

ความคิดเห็น • 6

  • @velugucharan8096
    @velugucharan8096 6 หลายเดือนก่อน +1

    It's correct

    • @fahdmirza
      @fahdmirza  6 หลายเดือนก่อน

      Thanks

  • @DEEPAKPAL01
    @DEEPAKPAL01 6 หลายเดือนก่อน

    Wow, that's great..
    I have a request. Can you please include Google Colab setup. This sholud help people who don't have locally available GPUs with them.

    • @fahdmirza
      @fahdmirza  6 หลายเดือนก่อน +1

      Sure thanks.

  • @OE-Vortex
    @OE-Vortex 6 หลายเดือนก่อน

    🎯 Key Takeaways for quick navigation:
    00:00 *Navarasa refers to the nine basic emotions fundamental to Human Experience in classical Indian art and literature, especially in the context of Bharatanatyam dance.*
    01:05 *Navarasa 2.0 is a Gemma instruction-tuned model available in 15 Indian languages and English, with two variants: 7 billion and 2 billion parameters.*
    01:44 *Navarasa 2.0 uses supervised fine-tuning (SFT) with a dataset specifying desired behavior for safety tuning.*
    02:39 *The video demonstrates how to locally install the Navarasa 2 billion model on Ubuntu, requiring the Transformers library.*
    04:46 *The model demonstrates translating English to Hindi and Urdu, showcasing its capability.*
    07:04 *Navarasa can also be installed and used on Windows using LM Studio, with demonstrations in multiple languages.*
    15:29 *Despite occasional inaccuracies, Navarasa shows promise for regional languages, with ongoing improvements expected.*
    Made with HARPA AI

    • @fahdmirza
      @fahdmirza  6 หลายเดือนก่อน

      cheers