meta-llama/llama3 on Hacker News

…open source'): https://opensource.org/blog/metas-llama-license-is-still-not... - A GitHub issue: https://github.com/meta-llama/llama3/issues/156 - A message from the the European Parliament about Meta's actions:…

comment by nailer · 2025-05-13 · in Meta's Llama license is not Open Source

…which, like Redis' SSPL, isn't, depite Meta saying it is), you might want to add your vote to: https://github.com/meta-llama/llama3/issues/156

comment by nailer · 2025-05-01 · in Redis is open source again

https://github.com/meta-llama/llama3/issues/156 if anyone else feels like commenting

comment by nailer · 2025-05-01 · in Everything we announced at our first LlamaCon

Llama 3 license: https://github.com/meta-llama/llama3/blob/main/LICENSE Calm it with the ad hominem attacks. It's not the place for it.

comment by caseyy · 2025-04-24 · in Careless People

…Here is the Llama source code, you can start training more epochs with it today if you like: https://github.com/meta-llama/llama3/blob/main/llama/model.p... It's rumored Llama 3 used FineWeb, but you're right that they at least haven't…

comment by causal · 2024-10-04 · in Meta Movie Gen

https://github.com/meta-llama/llama3/blob/main/llama/model.p...

comment by j_maffe · 2024-08-30 · in Update on Llama adoption

https://github.com/meta-llama/llama3 "To download the model weights and tokenizer, please visit the Meta Llama website and accept our License. Once your request is approved, you will receive a signed URL over email. Then, run the…

comment by arbven · 2024-08-29 · in Update on Llama adoption

…a clock rate of 1 ghz should result in a total time of 4 uSec/token, or 250,000 tokens/second. [1] https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md [2] https://mpedram.com/Papers/7nm-finfet-libraries-tcasII.pdf

comment by mikewarot · 2024-08-27 · in Cerebras Inference: AI at Instant Speed

…the opposite, you can modify the model, but you don't see all the details of its creation. Per https://github.com/meta-llama/llama3/blob/main/LICENSE there's also a laundry list of ways you're not allowed to use it, including restrictions…

comment by yjftsjthsd-h · 2024-07-23 · in Open source AI is the path forward

Llama’s code is open source: https://github.com/meta-llama/llama3/blob/main/llama/model.p...

comment by mkolodny · 2024-07-23 · in Open source AI is the path forward

…model code is shared (under llama license at least), which is really all you need to run training https://github.com/meta-llama/llama3/blob/main/llama/model.p...

comment by causal · 2024-07-23 · in Open source AI is the path forward

…any other large language model (excluding Meta Llama 3 or derivative works thereof). ... """ [0] https://github.com/meta-llama/llama3/blob/main/LICENSE

comment by abetusk · 2024-07-23 · in Open source AI is the path forward

…special token tokenization: Tokenization method with args to control special token handling: https://github.com/meta-llama/llama3/blob/bf8d18cd087a4a0b3f... And you can see how it is used combined with special tokens and user input here:…

comment by richdougherty · 2024-06-28 · in Gemma 2: Improving Open Language Models at a Practical Size [pdf]

…out that they were trained on publicly available datasets and NOT any Meta user data. For example: https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md

comment by jauer · 2024-06-13 · in How Meta trains large language models at scale

…is a bit different too: self.w2(F.silu(self.w1(x)) * self.w3(x)) I.e. the nonlinearity is a gate. https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959...

comment by jorlow · 2024-05-28 · in Reproducing GPT-2 in llm.c

I was curious so I tried to answer this question --- Training Llama 3 models emitted 2290 tons CO2e ( https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md ), and took 7.7 million GPU hours. Those GPU hours are for H100s, which…

comment by rhaps0dy · 2024-05-22 · in Hugging Face is sharing $10M worth of compute to help beat the big AI companies

…open source LLM fits in less than 300 lines of code and consists mostly of matrix multiplications. https://github.com/meta-llama/llama3/blob/main/llama/model.p... Anyone with a basic grasp of linear algebra can probably learn to understand…

comment by johndough · 2024-05-18 · in International Scientific Report on the Safety of Advanced AI [pdf]

…according to max_seq_len and max_batch_size values. So set those according to your hardware. [0] https://github.com/meta-llama/llama3/tree/14aab0428d3ec3a959...

comment by mkolodny · 2024-05-17 · in Llama 3 implemented in pure NumPy

Why is max_seq_len set to 2048 [1] when the model card says the context size is 8k [2]? [1] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959... [2] https://github.com/meta-llama/llama3/blob/14aab0428d3ec3a959...

comment by _pastel · 2024-05-17 · in Llama 3 implemented in pure NumPy

On line 59, there is a less-than-or-equals comparison between 0 and 1. Curious https://github.com/meta-llama/llama3/blob/main/llama/model.p...

comment by ebb_earl_co · 2024-05-16 · in Llama 3 implemented in pure NumPy

…that the original implementation by Meta is only 300 lines of very readable code [1]. [1]: https://github.com/meta-llama/llama3/blob/main/llama/model.p...

comment by ffriend · 2024-05-16 · in Llama 3 implemented in pure NumPy

https://github.com/meta-llama/llama3/commits?author=zuck turns up empty though. Maybe he just paid them

comment by rurban · 2024-04-23 · in Mark Zuckerberg himself appears in the list of direct contributors to Llama 3

▲ 3

Meta Llama 3 GitHub

story by adif_sgaid · 2024-04-22 · 0 comments

▲ 45

Mark Zuckerberg himself appears in the list of direct contributors to Llama 3

story by jonbraun · 2024-04-21 · 17 comments

…knowledge cutoff dates are March 2023 for the 8B version and December 2023 for the 70B version ( https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md )

comment by salesynerd · 2024-04-19 · in Meta Llama 3

…works thereof). """ [0] https://en.wikipedia.org/wiki/Open_source#Open-source_licens... [1] https://github.com/meta-llama/llama3/blob/main/LICENSE

comment by abetusk · 2024-04-19 · in Mark Zuckerberg: Llama 3, $10B Models, Caesar Augustus, Bioweapons [video]

…is pontificating about the benefits of them open sourcing their code when they haven't done so. [0] https://github.com/meta-llama/llama3 [1] https://en.wikipedia.org/wiki/Open_source

comment by abetusk · 2024-04-18 · in Mark Zuckerberg: Llama 3, $10B Models, Caesar Augustus, Bioweapons [video]

Yann Lecun's name is listed in the model card: https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md

comment by akgerber · 2024-04-18 · in Meta Llama 3

▲ 2

Llama 3 GitHub Repository

story by ushakov · 2024-04-18 · 0 comments

…to be the case at all. In fact I was surprised that Soumith Chintala was left out here [2]. [1] - https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md... [2] - https://soumith.ch/about/

comment by momofuku · 2024-04-18 · in Meta Llama 3

https://github.com/meta-llama/llama3/blob/main/LICENSE Llama is not open source. It's corporate freeware with some generous allowances. Open source licenses are a well defined thing. Meta marketing saying otherwise doesn't mean they get to…

comment by observationist · 2024-04-18 · in Meta Llama 3

…https://huggingface.co/meta-llama/Meta-Llama-3-70B https://llama.meta.com/llama-downloads https://github.com/meta-llama/llama3/blob/main/download.sh

comment by sroussey · 2024-04-18 · in Meta Llama 3

The model card has the benchmark results relative to other Llama models including Llama 2: https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md... The dramatic performance increase of Llama 3 relative to Llama 2 (even Llama 2 13B!)…

comment by minimaxir · 2024-04-18 · in Meta Llama 3