home.social

#bigcode — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #bigcode, aggregated by home.social.

  1. So that's me received the confirmation that my stuff is removed from Bigstack.

    Which is good. It shows the Optout requests are being done.

    Go to check again for librecasts old github account. Looks like I missed some.

    *Opens new ticket*

    huggingface.co/spaces/bigcode/

    While I did think it is important for Software Heritage to archive code, I wish it was done Opt-in.

    It would be nice to be asked and for that code to be curated. This is not curation. This is automation.

    #SoftwareHeritage #BigCode

  2. It’s especially rich that the logo for #BigCode, an org that trains LLMs so is massively accelerating #climateChange, uses a sakura blossom. Sakura are suddenly blooming earlier each year due to climate change.

  3. The Stack: 3 TB of permissively licensed source code

    Denis Kocetkov, Raymond Li, Loubna Ben allal et al.

    Action editor: Swarat Chaudhuri.

    openreview.net/forum?id=pxpbTd

    #bigcode #text2code #dataset

  4. #StarCoder: A State-of-the-Art #LLM for #Code by Hugging Face 🤗

    huggingface.co/blog/starcoder

    More about the Big Code project:
    bigcode-project.org/

    Find out, whether your code was used for training and opt-out, if you don't want to be "in the stack":
    huggingface.co/spaces/bigcode/

    #AI #ArtificialIntelligence #LLMs #DevTools #BigCode

  5. #StarCoder: A State-of-the-Art #LLM for #Code by Hugging Face 🤗

    huggingface.co/blog/starcoder

    More about the Big Code project:
    bigcode-project.org/

    Find out, whether your code was used for training and opt-out, if you don't want to be "in the stack":
    huggingface.co/spaces/bigcode/

    #AI #ArtificialIntelligence #LLMs #DevTools #BigCode

  6. #StarCoder: A State-of-the-Art #LLM for #Code by Hugging Face 🤗

    huggingface.co/blog/starcoder

    More about the Big Code project:
    bigcode-project.org/

    Find out, whether your code was used for training and opt-out, if you don't want to be "in the stack":
    huggingface.co/spaces/bigcode/

    #AI #ArtificialIntelligence #LLMs #DevTools #BigCode

  7. #StarCoder: A State-of-the-Art #LLM for #Code by Hugging Face 🤗

    huggingface.co/blog/starcoder

    More about the Big Code project:
    bigcode-project.org/

    Find out, whether your code was used for training and opt-out, if you don't want to be "in the stack":
    huggingface.co/spaces/bigcode/

    #AI #ArtificialIntelligence #LLMs #DevTools #BigCode

  8. I want to like HuggingChat open source LLM AI so much. But at least for coding it is nowhere near the same league as ChatGPT. If I would hire a new developer for my team and could conduct interviews only per keyboard, I would be impressed by ChatGPT and offer it the position. With HuggingChat I’d terminate the interview after 10 mins. Tried Java and JS. Calling APIs from any library that might do the job without importing, explanation mixing up cause and effect.. #chatgpt #huggingface #bigcode

  9. #BigCode is an open scientific collaboration working on responsible training of large language models for coding applications.

    In this organization you can find the artefacts of this collaboration:
    👉 #StarCoder, a state-of-the-art language model for code,
    👉 The #Stack, the largest available pretraining dataset with perimssive code, and 👉 #SantaCoder, a 1.1B parameter model for code.

    #StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages.
    It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle.

    Chat with StarCoder here: huggingface.co/chat/?model=big

    huggingface.co/bigcode

  10. #BigCode is an open scientific collaboration working on responsible training of large language models for coding applications.

    In this organization you can find the artefacts of this collaboration:
    👉 #StarCoder, a state-of-the-art language model for code,
    👉 The #Stack, the largest available pretraining dataset with perimssive code, and 👉 #SantaCoder, a 1.1B parameter model for code.

    #StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages.
    It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle.

    Chat with StarCoder here: huggingface.co/chat/?model=big

    huggingface.co/bigcode

  11. #BigCode is an open scientific collaboration working on responsible training of large language models for coding applications.

    In this organization you can find the artefacts of this collaboration:
    👉 #StarCoder, a state-of-the-art language model for code,
    👉 The #Stack, the largest available pretraining dataset with perimssive code, and 👉 #SantaCoder, a 1.1B parameter model for code.

    #StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages.
    It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle.

    Chat with StarCoder here: huggingface.co/chat/?model=big

    huggingface.co/bigcode

  12. #BigCode is an open scientific collaboration working on responsible training of large language models for coding applications.

    In this organization you can find the artefacts of this collaboration:
    👉 #StarCoder, a state-of-the-art language model for code,
    👉 The #Stack, the largest available pretraining dataset with perimssive code, and 👉 #SantaCoder, a 1.1B parameter model for code.

    #StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages.
    It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle.

    Chat with StarCoder here: huggingface.co/chat/?model=big

    huggingface.co/bigcode

  13. #BigCode is an open scientific collaboration working on responsible training of large language models for coding applications.

    In this organization you can find the artefacts of this collaboration:
    👉 #StarCoder, a state-of-the-art language model for code,
    👉 The #Stack, the largest available pretraining dataset with perimssive code, and 👉 #SantaCoder, a 1.1B parameter model for code.

    #StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages.
    It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle.

    Chat with StarCoder here: huggingface.co/chat/?model=big

    huggingface.co/bigcode

  14. #BigCode #OpenSource

    "#StarCoder is a 15.5B parameters language model for code trained for 1T tokens on 80+ programming languages. It uses MQA for efficient generation, has 8,192 tokens context window and can do fill-in-the-middle."

    huggingface.co/bigcode

  15. This one has different methodologies and philosophies that try to mitigate some of the ethical issues with other similar #GenerativeAI programming systems.

    Hugging Face and ServiceNow Research release StarCoder, a free alternative to code-generating #AI like GitHub's #Copilot, as part of the #BigCode project.

    techcrunch.com/2023/05/04/hugg

    #MachineLearning