Benefits	Disadvantages
Open model can be customized for your business.	Meta trained Llama 3 on public information, so, like other generative AI, it can create harmful content or blow up against copyright law.
Generate text or images, especially with the Imagine function for quick repetition of images.	In our tests, the professional emails generated by Llama 3 tended to be a little fuzzy and fluffy compared to those written by ChatGPT using GPT-3.5.
Use with your existing cloud provider.	As with other generative AI, training Llama 3 can be time and resource intensive if you intend to customize the model.

OpenAI may be the more familiar name when it comes to commercial generative AI, but Meta has successfully gained a foothold by open sourcing powerful large language models. Meta unveiled its largest generative AI model yet, Llama 3, on April 18, outperforming GPT-4 on some standard AI benchmark tests.

What is Llama 3?

Llama 3 is an LLM created by Meta. This can be used to create generative AI, including chatbots that can respond to a wide variety of queries in natural language. The use cases on which Llama 3 was evaluated include brainstorming ideas, creative writing, coding, summarizing documents and answering questions in the voice of a specific persona or character.

The complete Llama 3 model comes in four variants:

• 8 billion parameters pre-trained.
• 8 billion parameters instruction fine tuned.
• 70 billion parameters pre-trained.
• 70 billion parameters instruction fine tuned.

Llama 3’s generative AI capabilities can be used in a browser or through AI features in Meta’s Facebook, Instagram, WhatsApp and Messenger. The model itself can be downloaded from Meta or from major enterprise cloud platforms.

When will Llama 3 be released and on which platforms?

Llama 3 was released on April 18 on Google Cloud Vertex AI, IBM’s watsonx.ai and other major LLM hosting platforms. AWS followed suit and added Llama 3 to Amazon Bedrock on April 23rd. From April 29, Llama 3 is available on the following platforms:

• Databricks.
• Face hugging.
• Fireplace.
• Microsoft Azure.
• NVIDIA NIM.

Hardware platforms from AMD, AWS, Dell, Intel, NVIDIA and Qualcomm support Llama 3.

Llama 3.1

On July 23, Meta has announced Llama 3.1 405Bwhich is the most advanced version of Llama 3 yet, and improvements to Llama 3.1 70B and 8B. Llama 3.1B has 405 billion parameters, making it competitive with GPT-4o and Claude 3.5 Sonnet.

Is Llama 3 open source?

Llama 3 is open source, as were Meta’s other LLMs. Creating open source models has been a valuable differentiator for Meta. Meta founder and CEO Mark Zuckerberg explained his views on this philosophy in a July newsroom post.

SEE: Stanford’s AI Index report revealed 8 trends for AI in business today. (TechRepublic)

However, there is a debate about how much of a large language model’s code or weights must be publicly available to count as open source. Can Llama 3 really be called open source if there are some restrictions on its use, as there are for larger companies? Llama 3’s license agreement states that apps with more than 700 million monthly active users must request a separate license. But for business purposes, Meta offers a more open view of Llama 3 than its competitors do for their LLMs.

Is Llama 3 free?

Llama 3 is free as long as it is used under the terms of the license. The model can be downloaded directly from Meta or used within the various cloud hosting services listed above, although those services may have fees associated with them.

Is Llama 3 multimodal?

Llama 3 is not multimodal, which means that it is not able to understand data from different modalities such as video, audio or text. Meta plans to make Llama 3 multimodal in the near future.

Llama 3’s improvements over Llama 2

To make Llama 3 more capable than Llama 2, Meta added a new tokenizer to encode language much more efficiently. Meta created Llama 3 with clustered query attention, a method to improve the efficiency of model inferences. The Llama 3 training set is seven times the size of the training set used for Llama 2, Meta said, including four times as much code. Meta applied new efficiencies to Llama 3’s pre-training and instruction fine-tuning.

Since Llama 3 was designed as an open model, Meta added safety rails with developers in mind. A new guardrail is Code Shield, which is intended to trap insecure code that the model may produce.

Llama 3’s place in the competitive generative AI landscape

Llama 3 competes directly with GPT-4 and GPT-3.5, Google’s Gemini and Gemma, Mistral AI’s Mistral 7B, Perplexity AI and other LLMs for either individual or commercial use to build generative AI chatbots and other tools. About a week after Llama 3 was unveiled, Snowflake launched its own open enterprise AI with comparable capabilities called Snowflake Arctic.

The increasing performance requirements of LLMs like Llama 3 are contributing to an arms race of AI-enabled computers that can run models at least partially on the device. Meanwhile, generative AI companies may face increased scrutiny over heavy computing needs, which could contribute to exacerbating climate change.

Lama 3 vs GPT-4

Llama 3 outperforms OpenAI’s GPT-4 on HumanEval, which is a standard benchmark that compares the AI ​​model’s ability to generate code to code written by humans. Lama 3 70B scored 81.7, compared to GPT-4’s score of 67.

However, GPT-4 outperformed Llama 3 on the knowledge assessment MMLU with a score of 86.4 versus Llama 3 70B’s 79.5. Llama 3’s performance on more tests can be found at Meta’s blog post.

Should your organization use Llama 3?

Assuming you’ve already decided that generative AI is right for your business, choosing whether to use Llama 3 will likely depend on availability. Llama 3 is free to use and within limits more customizable than its competitors.

Llama 3 can be more effective than its competitors like GPT-4 or Claude 3 for encryption. However, Llama 3 has other competitors in the coding space. For example, GitHub recently debuted Copilot workspacewhich is adapted for coding and can create code based on natural language prompts.

Llama 3 might be good for your organization if you want a general purpose, open source family of AI models.