Hacker GitHub

← All repos

openlm-research/open_llama

View on GitHub ↗ ★ 7,517

OpenLLaMA, a permissively licensed open source reproduction of Meta AI’s LLaMA 7B trained on the RedPajama dataset

19 mentions on Hacker News by 14 people: 4 stories, 15 comments · first 2023-05-02 · latest 2023-07-11

Mentions per year 19 mentions · busiest year 2023 (19)
20082026
AllStoriesComments NewestTopOldest
  1. …into making it illegal for you to compete with us. https://github.com/ggerganov/llama.cpp https://github.com/openlm-research/open_llama https://huggingface.co/TheBloke/open-llama-7b-open-instruct-... https://huggingface.co/TheBloke/open-lla…
    comment by PostOnce · 2023-07-11 · in GPT-4 details leaked?
  2. OpenLLaMA is though. https://github.com/openlm-research/open_llama All of these are surmountable problems. We can beat OpenAI. We can drain their moat.
    comment by PostOnce · 2023-07-06 · in GPT-4 API General Availability
  3. …as well and is a straight drop-in (except for code generation due to multiple space tokenization: https://github.com/openlm-research/open_llama#update-0615202... ). XGen-7B is probably the superior 7B model, it's trained on more tokens and…
    comment by lhl · 2023-06-29 · in XGen-7B, a new 7B foundational model trained on up to 8K length for 1.5T tokens
  4. OpenLLaMA models up to 13B parameters have now been trained on 1T tokens: https://github.com/openlm-research/open_llama
    comment by jncraton · 2023-06-26 · in Databricks Strikes $1.3B Deal for Generative AI Startup MosaicML
  5. From their project page https://github.com/openlm-research/open_llama : > For current version of OpenLLaMA models, our tokenizer is trained to merge multiple empty spaces into one before tokenization, similar to T5 tokenizer. Because of…
    comment by loudmax · 2023-06-19 · in Open LLAMA 13B released, trained on 1T tokens
  6. …on how to have the LLaMa tokenizer (properly) handle repeating spaces, please see this discussion: https://github.com/openlm-research/open_llama/issues/40
    comment by hungrigekatze · 2023-06-18 · in OpenLLaMA 13B Released
  7. There is Open Llama 7B which is Apache 2.0 licensed, please consider checking it out https://github.com/openlm-research/open_llama
    comment by Kerbonut · 2023-06-15 · in Gorilla: Large Language Model Connected with APIs
  8. ▲ 1
    OpenLLaMA: Tokenizer Ignoring Multiple Spaces
    story by tosh · 2023-06-08 · 0 comments
  9. be sure to read the warning in their repo: https://github.com/openlm-research/open_llama#loading-the-we... > Please note that it is advised to avoid using the Hugging Face fast tokenizer for now, as we’ve observed that the auto-converted…
    comment by mdaniel · 2023-06-07 · in OpenLLaMA 7B Training Completed to 1T Tokens
  10. ▲ 2
    OpenLLaMA to train beyond 1T tokens
    story by tosh · 2023-06-07 · 0 comments
  11. OpenLLAMA will be released soon and it's 100% compatible with the original LLAMA. https://github.com/openlm-research/open_llama
    comment by digitallyfree · 2023-06-06 · in GGML – AI at the Edge
  12. …are already efforts to recreate the llama weights under open source licenses (eta: days/nowish). https://github.com/openlm-research/open_llama
    comment by tosh · 2023-05-26 · in Chatbot Arena Leaderboard
  13. …the evaluation results. We expect the full 1T token training run to finish at the end of this week. https://github.com/openlm-research/open_llama So we could develop on LLaMA for now and switch to OpenLLaMA later.
    comment by visarga · 2023-05-25 · in How to Finetune GPT-Like Large Language Models on a Custom Dataset
  14. …and well documented in the Github repo), there is also another group training "OpenLLaMA" https://github.com/openlm-research/open_llama (which looks like an effort by two Berkeley PhD students, https://www.haoliu.site/ and…
    comment by lhl · 2023-05-14 · in Open-Llama: Complete training pipeline for building large language models
  15. Mostly, yes. It's Apache License 2.0: https://github.com/openlm-research/open_llama/blob/main/LICE...
    comment by mkl · 2023-05-03 · in OpenLLaMA: An Open Reproduction of LLaMA
  16. …a lot of tasks[1] even though its trained on a fifth of the data (1 trillion vs 200 billion). [1] https://github.com/openlm-research/open_llama#evaluation
    comment by newswasboring · 2023-05-03 · in OpenLLaMA: An Open Reproduction of LLaMA
  17. At least for now they are focused on 7B and then 3B[1]. [1] https://github.com/openlm-research/open_llama#future-plans
    comment by newswasboring · 2023-05-03 · in OpenLLaMA: An Open Reproduction of LLaMA
  18. ▲ 484
    OpenLLaMA: An Open Reproduction of LLaMA
    story by sadiq · 2023-05-03 · 180 comments
  19. ▲ 4
    OpenLLaMA: An Open Reproduction of LLaMA
    story by Jimmc414 · 2023-05-02 · 0 comments