Difference between revisions of "LLaMA"

From
Jump to: navigation, search
m
m
Line 29: Line 29:
 
* [https://arstechnica.com/information-technology/2023/03/you-can-now-run-a-gpt-3-level-ai-model-on-your-laptop-phone-and-raspberry-pi/ You can now run a GPT-3 level AI model on your laptop, phone, and Raspberry Pi | Benj Edwards - Ars Technica]  ... On Friday, a software developer named Georgi Gerganov created a tool called [https://github.com/ggerganov/llama.cpp "llama.cpp"] that can run Meta's new GPT-3-class AI large language model, LLaMA, locally on a Mac laptop. Soon thereafter, people worked out how to run LLaMA on Windows as well. Then someone showed it running on a Pixel 6 phone, and next came a [https://www.raspberrypi.org/ Raspberry Pi] (albeit running very slowly).
 
* [https://arstechnica.com/information-technology/2023/03/you-can-now-run-a-gpt-3-level-ai-model-on-your-laptop-phone-and-raspberry-pi/ You can now run a GPT-3 level AI model on your laptop, phone, and Raspberry Pi | Benj Edwards - Ars Technica]  ... On Friday, a software developer named Georgi Gerganov created a tool called [https://github.com/ggerganov/llama.cpp "llama.cpp"] that can run Meta's new GPT-3-class AI large language model, LLaMA, locally on a Mac laptop. Soon thereafter, people worked out how to run LLaMA on Windows as well. Then someone showed it running on a Pixel 6 phone, and next came a [https://www.raspberrypi.org/ Raspberry Pi] (albeit running very slowly).
  
 +
LLaMA is a [[Large Language Model (LLM)]] released by Meta Platforms Inc. (formerly Facebook Inc.).
  
−
LLaMA is a [[Large Language Model (LLM)]] released by Meta Platforms Inc. (formerly Facebook Inc.) on February 24, 20231. Meta says it is democratizing access to LLMs, which are seen as one of the most important and beneficial forms of AI.  The four foundation models of LLaMA are LLaMA-7B, LLaMA-13B, LLaMA-33B, and LLaMA-65B12. They have 7 billion, 13 billion, 33 billion, and 65 billion parameters respectively. The models are all based on the [[transformer]] architecture and trained on publicly available datasets. LLaMA-13B is remarkable because it can run on a single GPU and outperform GPT-3 (175 billion parameters) on most common sense reasoning benchmarks. LLaMA-65B is competitive with the best models from other AI labs, such as Chinchilla 70B and PaLM 540B.
+
= LLaMA 2 =
 +
Meta released LLaMA in July 2023. The company said that it hopes that by making LLaMA open-source, it will be able to improve the model by getting feedback from the wider community of developers. LLaMA is still under development, but it has already been used to create some impressive chatbots. For example, one chatbot called Allie can be used to provide customer support. Another chatbot called Galactica is designed for scientific research.
 +
 
 +
= LLaMA =
 +
LLaMA on February 24, 20231. Meta says it is democratizing access to LLMs, which are seen as one of the most important and beneficial forms of AI.  The four foundation models of LLaMA are LLaMA-7B, LLaMA-13B, LLaMA-33B, and LLaMA-65B12. They have 7 billion, 13 billion, 33 billion, and 65 billion parameters respectively. The models are all based on the [[transformer]] architecture and trained on publicly available datasets. LLaMA-13B is remarkable because it can run on a single GPU and outperform GPT-3 (175 billion parameters) on most common sense reasoning benchmarks. LLaMA-65B is competitive with the best models from other AI labs, such as Chinchilla 70B and PaLM 540B.
  
  

Revision as of 21:05, 21 July 2023

YouTube search... ...Google search

LLaMA is a Large Language Model (LLM) released by Meta Platforms Inc. (formerly Facebook Inc.).

LLaMA 2

Meta released LLaMA in July 2023. The company said that it hopes that by making LLaMA open-source, it will be able to improve the model by getting feedback from the wider community of developers. LLaMA is still under development, but it has already been used to create some impressive chatbots. For example, one chatbot called Allie can be used to provide customer support. Another chatbot called Galactica is designed for scientific research.

LLaMA

LLaMA on February 24, 20231. Meta says it is democratizing access to LLMs, which are seen as one of the most important and beneficial forms of AI. The four foundation models of LLaMA are LLaMA-7B, LLaMA-13B, LLaMA-33B, and LLaMA-65B12. They have 7 billion, 13 billion, 33 billion, and 65 billion parameters respectively. The models are all based on the transformer architecture and trained on publicly available datasets. LLaMA-13B is remarkable because it can run on a single GPU and outperform GPT-3 (175 billion parameters) on most common sense reasoning benchmarks. LLaMA-65B is competitive with the best models from other AI labs, such as Chinchilla 70B and PaLM 540B.


Retrieved from "https://primo.ai/index.php?title=LLaMA&oldid=32371"