Power User
  • Communities
  • Create Post
  • Create Community
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
ashutrv@alien.topB to LocalLLaMAEnglish · 2 years ago

What are top open source projects in LLM space

message-square
message-square
13
link
fedilink
1
message-square

What are top open source projects in LLM space

ashutrv@alien.topB to LocalLLaMAEnglish · 2 years ago
message-square
13
link
fedilink

I am aware of langchain, LlamaIndex and MetaGPT https://github.com/geekan/MetaGPT https://github.com/langchain-ai/langchain https://github.com/run-llama/llama_index

alert-triangle
You must log in or register to comment.
  • 4onen@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    I’m unclear on your definition of “top” but

    • https://github.com/oobabooga/text-generation-webui - Popular open source llm frontend for just basic model usage.
    • https://github.com/ggerganov/llama.cpp - Popular llm backend (See: gguf file format.)
    • https://github.com/eth-sri/lmql - Best system for structured model outputs.
  • DecipheringAI@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    There’s also AutoGen, it’s quite popular on GitHub with almost 15k stars: https://github.com/microsoft/autogen

    It’s kinda similar to AutoGPT and it can also be used to generate code, like MetaGPT. I haven’t really tried it so I can’t say how good it is.

    • Feisty_Resolution157@alien.topB
      link
      fedilink
      English
      arrow-up
      1
      ·
      2 years ago

      The code generation is kind of a toy. It requires fairly specific outputs without even really doing that much to try and guide the outputs well. Cool, but very prototypy at best.

  • AsliReddington@alien.topB
    link
    fedilink
    arrow-up
    1
    ·
    2 years ago

    vLLM, TGI, Tensort-LLM

    Fuyu-8b

    • saintshing@alien.topB
      link
      fedilink
      English
      arrow-up
      1
      ·
      2 years ago

      mlc llm(deploying llms on mobiles and in browsers)

  • lakolda@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    LLMZip achieves SOTA compression by a large margin.

    • ashutrv@alien.topOPB
      link
      fedilink
      English
      arrow-up
      1
      ·
      2 years ago

      Very interesting 🤔 On what kind of data the compression is performing? Any progress in informational videos? I am working on building a smart media format too

      • lakolda@alien.topB
        link
        fedilink
        English
        arrow-up
        1
        ·
        2 years ago

        The original LLMZip paper mainly focused on text compression. A later work (I forget the name) used an LLM trained on byte tokens. This allowed it to compress not just text, but any file format. I think it may have been Google who published that particular paper… Very impressive though.

  • ksdio@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    https://github.com/bionic-gpt/bionic-gpt

    ​

    allows you to run a full LLM stack with RAG on your PC and also deploy to the cloud as the whole thing is dockerised

    A no code, Gen AI solution with RAG

  • ambient_temp_xeno@alien.topB
    link
    fedilink
    arrow-up
    1
    ·
    2 years ago

    I’m in spaaaaace!

  • 1ma_jones@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    You might want to check out the guidance DSL. It is conceptually similar to lmql but works on a lower level than lmql.

    https://github.com/guidance-ai/guidance/

    The library used to be developed by Microsoft. It seems like they spun it of as a independent company or nonprofit a few months ago.

  • Hey_You_Asked@alien.topB
    link
    fedilink
    arrow-up
    1
    ·
    2 years ago

    RemindMe! 2 weeks

  • ErikBjare@alien.topB
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 years ago

    I keep an up-to-date list of code generation and Copilot-like tools here: https://github.com/ErikBjare/are-copilots-local-yet

LocalLLaMA

localllama

Subscribe from Remote Instance

Create a post
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !localllama@poweruser.forum

Community to discuss about Llama, the family of large language models created by Meta AI.

Visibility: Public
globe

This community can be federated to other instances and be posted/commented in by their users.

  • 4 users / day
  • 4 users / week
  • 4 users / month
  • 4 users / 6 months
  • 1 local subscriber
  • 11 subscribers
  • 1.02K Posts
  • 5.82K Comments
  • Modlog
  • mods:
  • communick
  • BE: 0.19.11
  • Modlog
  • Instances
  • Docs
  • Code
  • join-lemmy.org