Это видео недоступно.
Сожалеем об этом.

Exploring StarCoder: Open Source LLM for Code Completion

Поделиться
HTML-код
  • Опубликовано: 8 май 2023
  • StarCoder, the hottest new Open Source code-completion LLM, is based on GPT-2 architecture and trained on The Stack - which contains an insane amount of permissive code.
    Star Coder shows how open source AI is advancing fast. The model may not match GPT-4 but it highlights how the community is gaining capabilities that are on pace to match industry titans such as Google and OpenAI!
    Overall, Star Coder is a solid model for code completion!
    Check it out - it's totally open source!
    🔗Blog: huggingface.co/blog/starcoder
    🔗Paper: drive.google.com/file/d/1cN-b...
    🔗huggingface.co/chat/
    #llm #generativeai #huggingface #starcoder

Комментарии • 16

  • @OlabodeAdedoyin
    @OlabodeAdedoyin Год назад +6

    Please, don't stop making these

  • @creativeuser9086
    @creativeuser9086 Год назад +3

    Can you do a review on all of the other code copilots and show us which is the best?

    • @chrisalexiuk
      @chrisalexiuk  Год назад +2

      I might be able to do that in the coming weeks, sure! I want to wait for GitHub to announce their latest tooling before I do, though!

  • @user-lg5bm1ju8h
    @user-lg5bm1ju8h Год назад +1

    Very informative video, thank you.
    One question about HuggingFace Chat. How did you get the change model option ? I see the current chat version is v0.2 and in this version or in my account at least, the change model option is missing.
    Thank you

    • @chrisalexiuk
      @chrisalexiuk  Год назад

      It looks like they've removed that particular option from the Chat demo!
      You could still use the model by heading a Space powered by the Chat model!

  • @Machiuka
    @Machiuka 6 месяцев назад

    How to use it locally through gpt4all? Thank you!

  • @amparoconsuelo9451
    @amparoconsuelo9451 11 месяцев назад

    Can a subsequent SFT and RTHF with different, additional or lesser contents change the character, improve, or degrade a GPT model? Can you modify a GPT model?

  • @arvj123
    @arvj123 Год назад +1

    I was so looking forward to using this LLM to use on my iOS project and then the paper said, "Out of the languages present in MultiPL-E (Cassano et al., 2023), only D and Swift were not included in the training set. For D, language misclassification of the files led to less than 2MB of data in The Stack (Kocetkov et al., 2022). For Swift, due to a human mistake, it was not included in the final list of languages." lmao wat

    • @chrisalexiuk
      @chrisalexiuk  Год назад +2

      That is.... unfortunate.
      Time to create your own set!

  • @justinmilner8
    @justinmilner8 Год назад +1

    Anyone know if it's possible to host Star Coder locally?

    • @chrisalexiuk
      @chrisalexiuk  Год назад

      Yes! Definitely!
      You'll need some decent hardware - but it's definitely possible.
      Checkout the docs!

    • @justinmilner8
      @justinmilner8 Год назад +1

      @@chrisalexiuk Thanks Chris, looks like they've made it pretty easy to configure HF Code Autocomplete to point to custom endpoints, awesome.

  • @HaseebHeaven
    @HaseebHeaven Год назад +2

    This needs lots of improvement the Codex AI is far better model

    • @chrisalexiuk
      @chrisalexiuk  Год назад +3

      Definitely true! While it's better than the original Codex model (in metrics) it has a lot of distance to cover before it can exceed the newest Codex model.

  • @raymondfloyd6397
    @raymondfloyd6397 Год назад

    😌 "Promosm"