presentation mode on windows

At the edge of tweaking

Advertisement

Command Line Options

The Presentation Mode feature is implemented via an executable file, presentationsettings.exe .

By simply running it, you will open the Settings dialog shown above.

It supports the following command line options:

This command will directly enable Presentation Mode.

The next command will disable it:

Winaero greatly relies on your support. You can help the site keep bringing you interesting and useful content and software by using these options:

If you like this article, please share it using the buttons below. It won't take a lot from you, but it will help us grow. Thanks for your support!

Author: Sergey Tkachenko

Sergey Tkachenko is a software developer who started Winaero back in 2011. On this blog, Sergey is writing about everything connected to Microsoft, Windows and popular software. Follow him on Telegram , Twitter , and YouTube . View all posts by Sergey Tkachenko

4 thoughts on “Enable Presentation Mode in Windows 10”

Does this also prevent reboots from Windows 10 crappy updates?

I doubt that

I generally use the “pause updates” feature the day before I give a presentation. It should stop windows doing any updates/reboots while you’re delivering a talk :)

Sadly, it seems that the presentationsettings.exe is only available on Windows 10 PRO, not on home.

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

Privacy Overview

How to Set Presentation Mode in Windows 11

How do you set presentation mode in windows 11.

Presentation mode in Windows 11 is a very useful way to ensure the user experience is the best possible. Helps keep programs, applications and files organized easily for best results. If you want to set up presentation mode for Windows 11, follow these simple steps.

Step 1: Open Control Panel

To access the control panel, open the Start menu and click the Control Panel icon.

Step 2: Adjust the parameters

At the top of the Control Panel window there is an option called Presentation Settings, click on it. In this window you can modify taskbar settings, desktop settings, thumbnails, and more.

Step 3: Customize the screen

To customize the screen to make it look better, click the Customize Screen option at the bottom of the window. This option will give you a number of options to choose from, such as the screen background theme, color layout, and thumbnail size.

Step 4: Save the changes.

Once you have made the desired changes, remember to click the Save Changes button so that the changes are saved. This way, the settings will remain saved even when you restart your computer.

Examples of Presentation Mode:

• Wide taskbar : This option will add a second row to the taskbar to allow you to view more apps at the same time.
• Large desk folders : This option will set all desktop folders and files to a larger thumbnail size so you can see them in greater detail.
• Transparent miniatures : This option makes app and file thumbnails easier to identify by displaying them as a transparent background.

Recommendation:

It is always recommended that users configure presentation mode every time they install Windows 11, as this will make the system better suited to their needs and easier to use.

If you want to change the settings at any time, you can always reopen the Control Panel and make the necessary changes.

Presentation mode settings in Windows 11

Windows 11 includes specific presentation modes to help users beautifully present their content. This feature offers many useful options for setting aspects such as the appearance and behavior of the projector that allow you to optimize your presentations. Below we show you how to configure presentation mode in Windows 11.

Step 1: Open the Settings Menu

Setting up presentation mode in Windows 11 begins by opening the Settings Menu. To do this, press the key Windows + I Next, click on System and then on the Projection section.

Step 2: Set Presentation Mode

Now, click on the Presentation Mode button to open a new window with configuration options.

In this window there are several possible configurations:

• Design: here you can set the appearance of what is displayed on the screen.
• Advanced: In this section there are options related to the behavior of the projector. For example, you can configure whether you want the toolbar to be automatically displayed on the screen or whether you want it to be hidden.
• Background: From here you can set a background for your presentation. You can choose between a solid color or a default image.

Once you're done, click the Apply button to save your changes.

Example: Setting up a Presentation Background

As an example, we are going to set up a presentation background with an image. To do this, open the presentation mode settings window and click on the Background section. Select the Image radius and click the Browse button to find the image you want to use. Next, click the Apply button to save the changes and you will see your presentation background ready to use.

Step 3: Save your Presentation

Once you've finished setting up your projector, it's time to save the presentation to make the changes permanent. To do this, return to the presentation mode settings window and click the Save button to save the changes.

The process to configure the presentation mode in Windows 11 is simple and fast, with these simple steps you can optimize your content to present it professionally.

You may also be interested in this related content:

• How to Update Whatsapp on Android
• How to Login to Messenger
• How to Get the New Emojis on iPhone

Stack Exchange Network

Stack Exchange network consists of 183 Q&A communities including Stack Overflow , the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Q&A for work

Connect and share knowledge within a single location that is structured and easy to search.

Enable Presentation Mode permanently on Windows

I want to enable the presentation mode forever on my Windows 10 device. When I turn it on, it again turns off after reboot or sign out. I have also tried to set the values through registry, but seems not to be working.

HKEY_CURRENT_USER\Software\Microsoft\MobilePC\AdaptableSettings\SkipBatteryCheck > 1

HKEY_CURRENT_USER\Software\Microsoft\MobilePC\MobilityCenter\RunOnDesktop > 1

The main reason why I want to enable it is, it's an unattended PC and I have tried all power settings to turn the sleep off. But after some time it goes to sleep mode.

• power-management

• So, from what I understand, you don't need to enable windows presentation mode, you need to prevent the PC from sleeping right? –  Genesis_GDK Jun 6, 2017 at 13:45
• Yes. Since I tried all other option that I know and those are not working. But this presentation mode works well unless the device is rebooted. –  Gaurav Jun 6, 2017 at 14:52

2 Answers 2

In order to enable presentation mode forever, I suggest you create a startup script to invoke it.

To control the presentation settings, you can call the executable directly, followed by the arguments. Example:

or, to stop:

If the first line is scheduled to run at boottime then Presentation mode will always be activated.

• Can you confirm that this works for you? –  Genesis_GDK Jun 11, 2017 at 11:24

Your main goal from what your question shows, is to prevent your system from switching off the screen or hibernating or anything like that. In which case, if you have tried disabling all the energy saving features that you know of, and the PC is still hibernating, then I suggest you use Caffeine. It costs nothing and it's adware free.

Caffeine works in a rather simple way to try and prevent your computer going into standby, the screen from shutting off or even the screensaver kicking in. This is to simulate a key press once every 59 seconds so Windows is fooled into thinking you’re using your computer when in fact, you aren’t.

Get it from zhornsoftware, and the link below: zhornsoftware.co.uk/caffeine/index.html#download

• Yes, I have replicated same thing in my application (which runs always on device) what Caffeine does, triggering F15 key press event every 50 seconds. Will have to wait till 24 hours to make sure that it's working. Isn't there a way to enable the presentation mode forever? –  Gaurav Jun 6, 2017 at 14:55

You must log in to answer this question.

Not the answer you're looking for browse other questions tagged windows windows-10 sleep power-management iot ..

• The Overflow Blog
• Supporting the world’s most-used database engine through 2050
• What language should beginning programmers choose?

Hot Network Questions

• Integral of the product of a Gaussian and a exponential of a hyperbolic function
• What does modern physics hold to be the “base ontological system” of reality?
• Enumerate all matches of a regex
• Using DeleteDuplicates with choosing which item of the pair should be deleted
• Conditions for Binomial Distribution
• Taking a scene from a video I made
• Does washing the bike (not the drivetrain) actually prolong the life of any components?
• Ways to refuel nuclear powered cars
• Shebang can reference a script in Linux
• Proving that a function is one-to-one (injective)
• Help needed for Lean4 book Interlude exercise
• How to make a sign language that only uses three fingers
• Can copy-pasting a word definition from a dictionary site cause a copyright issue?
• Why did I lose a point of rating in stalemate?
• When submitting a letter to an agent for publication of a novel, what does it mean to say that your manuscript is "complete"?
• Does a WIPO patent supersede a US patent?
• Book series about a guy who wins the lottery and builds an elaborate post-apocalyptic bunker
• How was the miraculous complement of two Harkonnen legions for the retaking of Arrakis explained?
• Infer pluses and minuses
• Unable to detach kitchen faucet
• Reference for a statement from Gaitsgory's thesis
• Diagonal Cracks at corner of 1 piece shower/tub - Serious?
• Alternate history - Blondlot's N-Rays really exist, leading to French dominance in 20th century
• Why might a flight in turbulence turn off cabin lights and heating?

How-To Geek

How to control what multiple displays do on windows 10.

Gain control of your multiple-monitor Windows 10 setup with a simple keyboard shortcut.

Quick Links

Using the windows + p "project" menu, examples of when to use each project option, more multi-display tips.

Windows can use multiple displays in several ways, extending or duplicating your primary monitor. You can even turn off your primary monitor. Here's how to change secondary display settings with a simple keyboard shortcut.

To quickly change how Windows 10 handles multiple displays, press Windows + P.

A dark-gray menu titled "Project" will pop up on the right part of your screen. (That's "project" as in projector, not "project" as in a task people work on together.)

From here, you can use the mouse or cursor keys and Enter to choose the option you'd like. Here's what the options do:

• PC screen only : This mode only shows video on the primary display, even if a second one is connected.
• Duplicate: This mode duplicates the video output from the primary display onto a second display, which can be ideal for presentations.
• Extend: This mode extends the desktop across as many displays as you have connected. It stitches them all into one large virtual desktop, and you can move windows between them .
• Second screen only: This mode only shows video on the second display, while the primary display remains blank and unused.

Related: How to Rearrange Windows with Keyboard Shortcuts on Windows 10

With the options listed above in mind, let's take a look at four common multiple display scenarios and examine how these options can help:

Scenario 1: PowerPoint Presentation

You have a laptop connected to a digital projector in a lecture hall, business meeting, or church, and you'd like to display the contents of your screen through the projector. In this case, you'd want to choose Duplicate mode from the Windows+P Project menu. Next, start up your presentation program as usual, and the audience will see exactly what you see on your computer screen.

Scenario 2: Desktop Power User

You have two or more monitors to give you more screen real estate for productivity purposes, such as software development, video editing, stock trading, or music production. In this case, you'd want to use the Extend mode from the Windows+P Project menu, so you can see as much information as possible at once. You can even move windows between monitors with a keyboard shortcut .

Scenario 3: Laptop Monitor Replacement

You have a work laptop that you'd like to use at home with a large monitor, but you don't need the laptop's built-in display. You can even configure the laptop to stay running with the lid closed, and use an external keyboard and a mouse. In this case, you'd want to pick Second screen only mode from the Windows + P Project menu.

Scenario 4: Family Movie Night

It's movie night, and you have a new video projector that you'd like to use to watch a film from your PC. In this case, you can use Duplicate or Second screen only mode from the Windows + P Project menu. If the projector has a different native resolution than your display, then "Second screen only" may work better because Windows will adapt the output perfectly to the projector instead of trying to stretch your primary display's resolution to fit.

Related: How to Move a Window to Another Monitor on Windows 10

If you don't have a spare monitor but have a PC with a built-in display, you can potentially use it as second monitor with a wireless technology called Miracast built into Windows 10. And, for more info on working productively with multiple-monitor setups, check out our in-depth guides to taking advantage of dual monitors  and adding a second screen to laptops . Have fun, and enjoy the view!

Related: How to Use Multiple Monitors to Be More Productive

• Use Presenter View Video
• Add speaker notes​ Video
• Practice and time your presentation​ Video
• Record a presentation Video
• Print a presentation Video

Use Presenter View

PowerPoint Presenter View shows you the current slide, the next slide, and your speaker notes, to help you focus while presenting and connect with your audience. 

Select the Slide Show tab.

Select the Use Presenter View checkbox.

Select which monitor to display Presenter View on.

In Presenter View , you can:

See your current slide, next slide, and speaker notes.

Select the arrows next to the slide number to go between slides.

Select the pause button or reset button to pause or reset the slide timer in the upper left.

See the current time to help you pace your presentation.

Select the font icons to make the speaker notes larger or smaller.

Select the annotations pen icon to draw on the screen in real time, or select a laser pointer.

Select the thumbnail icon to see all the slides in your presentation and quickly jump to another slide.

Select the magnifying glass icon to zoom in on a particular part of a slide.

The screen icon let's you make the screen temporarily black to focus the attention on you.

Select END SLIDESHOW when you're done presenting.

Start the presentation and see your notes in Presenter view

Need more help?

Want more options.

Explore subscription benefits, browse training courses, learn how to secure your device, and more.

Microsoft 365 subscription benefits

Microsoft 365 training

Microsoft security

Accessibility center

Communities help you ask and answer questions, give feedback, and hear from experts with rich knowledge.

Ask the Microsoft Community

Microsoft Tech Community

Windows Insiders

Microsoft 365 Insiders

Was this information helpful?

Thank you for your feedback.

Contribute to the Windows forum! Click  here  to learn more  💡

April 9, 2024

Contribute to the Windows forum!

Click  here  to learn more  💡

Windows 10 Top Forum Contributors: neilpzz  -  Ramesh Srinivasan  -  _AW_  -  questions_  -  Volume Z 👍✅

April 17, 2024

Windows 10 Top Forum Contributors:

neilpzz  -  Ramesh Srinivasan  -  _AW_  -  questions_  -  Volume Z 👍✅

• Search the community and support articles
• Search Community member

Ask a new question

Presentation setttings missing

When I open the Mobility Center, it offers options to set brightness, volume, battery, external display, and sync, but not presentation settings.

How do I turn that on?

Report abuse

Replies (4) .

• Independent Advisor

1 person found this reply helpful

Was this reply helpful? Yes No

Sorry this didn't help.

Great! Thanks for your feedback.

How satisfied are you with this reply?

Thanks for your feedback, it helps us improve the site.

Thanks for your feedback.

No, I know how to do that.

Under Windows + X ==> Mobility Center

there is a set of tiles for choosing various things. It's supposed to have a presentation settings feature, where I can turn off certain things that would interrupt a presentation. I do a lot of that, and find the various announcements, etc. more than a little annoying.

See the discussion (and screenshots) at

https://www.thewindowsclub.com/presentation-settings-in-windows-7/

I am missing the last tile, where I could turn on or off those annoyances.

Is this something that was in Win 7 and eliminated for Win 10?

Yep...just figured that out...I have "Home and Business"

Guess I need to look into an upgrade, if I want it to do that.

2 people found this reply helpful

Question Info

• Norsk Bokmål
• Ελληνικά
• Русский
• עברית
• العربية
• ไทย
• 한국어
• 中文(简体)
• 中文(繁體)
• 日本語

This browser is no longer supported.

Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.

View reports and dashboards in presentation mode on Windows devices

• 9 contributors

The Power BI Windows app was retired on December 31st, 2023, and is no longer supported. It's been removed from the Microsoft apps store, and no further product updates for it will be released. Please use the Power BI service to view your Power BI content on Windows devices.

This notice only applies to the Power BI Windows app. It does not affect the Power BI Desktop app.

You can use presentation mode to display reports and dashboards in full-screen on Windows devices. Presentation mode is useful for displaying Power BI at meetings or conferences, or on a dedicated projector at the office, or even just for maximizing space on a small screen.

In presentation mode:

• All the "chrome" (such as the navigation and menu bars) disappears, making it easier to focus on the data in your report.
• An action toolbar becomes available to enable you to interact with your data and to control the presentation.
• You can play a slideshow that cycles automatically between pages, bookmarks, or both pages and bookmarks.

Use presentation mode

From the toolbar you can tap to perform the following actions:

You can undock the toolbar and drag and drop it to anywhere on the screen. This is useful for large screens, when you want to focus on a specific area in your report and want to have the tools available next to it. Just place your finger on the toolbar and swipe it into the report canvas.

You can play a slideshow to automatically cycle through your presentation. You can set the slideshow to cycle through pages, bookmarks, or both pages and bookmarks. During the slideshow, report pages with auto page refresh will continue to automatically refresh as configured, ensuring that the most current data is always shown.

When you select the Play button on the action toolbar, the slideshow begins. A controller appears that allows you to pause the slideshow or to change what's being played: pages, bookmarks, or both pages and bookmarks.

The controller shows the name of the currently displayed view (page or bookmark and page). In the image above, we see that in the report called Sales , we are currently viewing the Asia Pacific bookmark on the Sales Performance page.

By default, a slideshow cycles through pages only, at a rate of one every 30 seconds. You can change the default behavior in the Slideshow settings .

Auto play a slideshow on startup

You can set up the Power BI Windows app to start playing a slideshow automatically whenever the app is launched. This option is useful for creating a kiosk-like experience that runs a report in public displays without any manual intervention. See Slideshow settings for detail about setting up a report for autoplay.

Slideshow settings

By default, a slideshow cycles through pages only, at a rate of one every 30 seconds. You can change this default behavior by going to Settings > Options , as illustrated below. You can also turn on autoplay and choose a report to play.

Select the settings icon.

Open the options page.

If desired, change the default settings for what the slideshow will cycle over (pages, bookmarks, or both) and how frequently the slides will transition.

If you want your report to start playing automatically when the app is launched, turn on the toggle and choose Select Report . You will be able to search for reports you have access to.

Related content

• Display dashboards and reports in full-screen mode from the Power BI service
• Questions? Try asking the Power BI Community

Was this page helpful?

Coming soon: Throughout 2024 we will be phasing out GitHub Issues as the feedback mechanism for content and replacing it with a new feedback system. For more information see: https://aka.ms/ContentUserFeedback .

Submit and view feedback for

Additional resources

Our approach

• Responsibility
• Infrastructure
• Try Meta AI

RECOMMENDED READS

• 5 Steps to Getting Started with Llama 2
• The Llama Ecosystem: Past, Present, and Future
• Introducing Code Llama, a state-of-the-art large language model for coding
• Meta and Microsoft Introduce the Next Generation of Llama
• Today, we’re introducing Meta Llama 3, the next generation of our state-of-the-art open source large language model.
• Llama 3 models will soon be available on AWS, Databricks, Google Cloud, Hugging Face, Kaggle, IBM WatsonX, Microsoft Azure, NVIDIA NIM, and Snowflake, and with support from hardware platforms offered by AMD, AWS, Dell, Intel, NVIDIA, and Qualcomm.
• We’re dedicated to developing Llama 3 in a responsible way, and we’re offering various resources to help others use it responsibly as well. This includes introducing new trust and safety tools with Llama Guard 2, Code Shield, and CyberSec Eval 2.
• In the coming months, we expect to introduce new capabilities, longer context windows, additional model sizes, and enhanced performance, and we’ll share the Llama 3 research paper.
• Meta AI, built with Llama 3 technology, is now one of the world’s leading AI assistants that can boost your intelligence and lighten your load—helping you learn, get things done, create content, and connect to make the most out of every moment. You can try Meta AI here .

Today, we’re excited to share the first two models of the next generation of Llama, Meta Llama 3, available for broad use. This release features pretrained and instruction-fine-tuned language models with 8B and 70B parameters that can support a broad range of use cases. This next generation of Llama demonstrates state-of-the-art performance on a wide range of industry benchmarks and offers new capabilities, including improved reasoning. We believe these are the best open source models of their class, period. In support of our longstanding open approach, we’re putting Llama 3 in the hands of the community. We want to kickstart the next wave of innovation in AI across the stack—from applications to developer tools to evals to inference optimizations and more. We can’t wait to see what you build and look forward to your feedback.

Our goals for Llama 3

With Llama 3, we set out to build the best open models that are on par with the best proprietary models available today. We wanted to address developer feedback to increase the overall helpfulness of Llama 3 and are doing so while continuing to play a leading role on responsible use and deployment of LLMs. We are embracing the open source ethos of releasing early and often to enable the community to get access to these models while they are still in development. The text-based models we are releasing today are the first in the Llama 3 collection of models. Our goal in the near future is to make Llama 3 multilingual and multimodal, have longer context, and continue to improve overall performance across core LLM capabilities such as reasoning and coding.

State-of-the-art performance

Our new 8B and 70B parameter Llama 3 models are a major leap over Llama 2 and establish a new state-of-the-art for LLM models at those scales. Thanks to improvements in pretraining and post-training, our pretrained and instruction-fine-tuned models are the best models existing today at the 8B and 70B parameter scale. Improvements in our post-training procedures substantially reduced false refusal rates, improved alignment, and increased diversity in model responses. We also saw greatly improved capabilities like reasoning, code generation, and instruction following making Llama 3 more steerable.

*Please see evaluation details for setting and parameters with which these evaluations are calculated.

In the development of Llama 3, we looked at model performance on standard benchmarks and also sought to optimize for performance for real-world scenarios. To this end, we developed a new high-quality human evaluation set. This evaluation set contains 1,800 prompts that cover 12 key use cases: asking for advice, brainstorming, classification, closed question answering, coding, creative writing, extraction, inhabiting a character/persona, open question answering, reasoning, rewriting, and summarization. To prevent accidental overfitting of our models on this evaluation set, even our own modeling teams do not have access to it. The chart below shows aggregated results of our human evaluations across of these categories and prompts against Claude Sonnet, Mistral Medium, and GPT-3.5.

Preference rankings by human annotators based on this evaluation set highlight the strong performance of our 70B instruction-following model compared to competing models of comparable size in real-world scenarios.

Our pretrained model also establishes a new state-of-the-art for LLM models at those scales.

To develop a great language model, we believe it’s important to innovate, scale, and optimize for simplicity. We adopted this design philosophy throughout the Llama 3 project with a focus on four key ingredients: the model architecture, the pretraining data, scaling up pretraining, and instruction fine-tuning.

Model architecture

In line with our design philosophy, we opted for a relatively standard decoder-only transformer architecture in Llama 3. Compared to Llama 2, we made several key improvements. Llama 3 uses a tokenizer with a vocabulary of 128K tokens that encodes language much more efficiently, which leads to substantially improved model performance. To improve the inference efficiency of Llama 3 models, we’ve adopted grouped query attention (GQA) across both the 8B and 70B sizes. We trained the models on sequences of 8,192 tokens, using a mask to ensure self-attention does not cross document boundaries.

Training data

To train the best language model, the curation of a large, high-quality training dataset is paramount. In line with our design principles, we invested heavily in pretraining data. Llama 3 is pretrained on over 15T tokens that were all collected from publicly available sources. Our training dataset is seven times larger than that used for Llama 2, and it includes four times more code. To prepare for upcoming multilingual use cases, over 5% of the Llama 3 pretraining dataset consists of high-quality non-English data that covers over 30 languages. However, we do not expect the same level of performance in these languages as in English.

To ensure Llama 3 is trained on data of the highest quality, we developed a series of data-filtering pipelines. These pipelines include using heuristic filters, NSFW filters, semantic deduplication approaches, and text classifiers to predict data quality. We found that previous generations of Llama are surprisingly good at identifying high-quality data, hence we used Llama 2 to generate the training data for the text-quality classifiers that are powering Llama 3.

We also performed extensive experiments to evaluate the best ways of mixing data from different sources in our final pretraining dataset. These experiments enabled us to select a data mix that ensures that Llama 3 performs well across use cases including trivia questions, STEM, coding, historical knowledge, etc.

Scaling up pretraining

To effectively leverage our pretraining data in Llama 3 models, we put substantial effort into scaling up pretraining. Specifically, we have developed a series of detailed scaling laws for downstream benchmark evaluations. These scaling laws enable us to select an optimal data mix and to make informed decisions on how to best use our training compute. Importantly, scaling laws allow us to predict the performance of our largest models on key tasks (for example, code generation as evaluated on the HumanEval benchmark—see above) before we actually train the models. This helps us ensure strong performance of our final models across a variety of use cases and capabilities.

We made several new observations on scaling behavior during the development of Llama 3. For example, while the Chinchilla-optimal amount of training compute for an 8B parameter model corresponds to ~200B tokens, we found that model performance continues to improve even after the model is trained on two orders of magnitude more data. Both our 8B and 70B parameter models continued to improve log-linearly after we trained them on up to 15T tokens. Larger models can match the performance of these smaller models with less training compute, but smaller models are generally preferred because they are much more efficient during inference.

To train our largest Llama 3 models, we combined three types of parallelization: data parallelization, model parallelization, and pipeline parallelization. Our most efficient implementation achieves a compute utilization of over 400 TFLOPS per GPU when trained on 16K GPUs simultaneously. We performed training runs on two custom-built 24K GPU clusters . To maximize GPU uptime, we developed an advanced new training stack that automates error detection, handling, and maintenance. We also greatly improved our hardware reliability and detection mechanisms for silent data corruption, and we developed new scalable storage systems that reduce overheads of checkpointing and rollback. Those improvements resulted in an overall effective training time of more than 95%. Combined, these improvements increased the efficiency of Llama 3 training by ~three times compared to Llama 2.

Instruction fine-tuning

To fully unlock the potential of our pretrained models in chat use cases, we innovated on our approach to instruction-tuning as well. Our approach to post-training is a combination of supervised fine-tuning (SFT), rejection sampling, proximal policy optimization (PPO), and direct preference optimization (DPO). The quality of the prompts that are used in SFT and the preference rankings that are used in PPO and DPO has an outsized influence on the performance of aligned models. Some of our biggest improvements in model quality came from carefully curating this data and performing multiple rounds of quality assurance on annotations provided by human annotators.

Learning from preference rankings via PPO and DPO also greatly improved the performance of Llama 3 on reasoning and coding tasks. We found that if you ask a model a reasoning question that it struggles to answer, the model will sometimes produce the right reasoning trace: The model knows how to produce the right answer, but it does not know how to select it. Training on preference rankings enables the model to learn how to select it.

Building with Llama 3

Our vision is to enable developers to customize Llama 3 to support relevant use cases and to make it easier to adopt best practices and improve the open ecosystem. With this release, we’re providing new trust and safety tools including updated components with both Llama Guard 2 and Cybersec Eval 2, and the introduction of Code Shield—an inference time guardrail for filtering insecure code produced by LLMs.

We’ve also co-developed Llama 3 with torchtune , the new PyTorch-native library for easily authoring, fine-tuning, and experimenting with LLMs. torchtune provides memory efficient and hackable training recipes written entirely in PyTorch. The library is integrated with popular platforms such as Hugging Face, Weights & Biases, and EleutherAI and even supports Executorch for enabling efficient inference to be run on a wide variety of mobile and edge devices. For everything from prompt engineering to using Llama 3 with LangChain we have a comprehensive getting started guide and takes you from downloading Llama 3 all the way to deployment at scale within your generative AI application.

A system-level approach to responsibility

We have designed Llama 3 models to be maximally helpful while ensuring an industry leading approach to responsibly deploying them. To achieve this, we have adopted a new, system-level approach to the responsible development and deployment of Llama. We envision Llama models as part of a broader system that puts the developer in the driver’s seat. Llama models will serve as a foundational piece of a system that developers design with their unique end goals in mind.

Instruction fine-tuning also plays a major role in ensuring the safety of our models. Our instruction-fine-tuned models have been red-teamed (tested) for safety through internal and external efforts. ​​Our red teaming approach leverages human experts and automation methods to generate adversarial prompts that try to elicit problematic responses. For instance, we apply comprehensive testing to assess risks of misuse related to Chemical, Biological, Cyber Security, and other risk areas. All of these efforts are iterative and used to inform safety fine-tuning of the models being released. You can read more about our efforts in the model card .

Llama Guard models are meant to be a foundation for prompt and response safety and can easily be fine-tuned to create a new taxonomy depending on application needs. As a starting point, the new Llama Guard 2 uses the recently announced MLCommons taxonomy, in an effort to support the emergence of industry standards in this important area. Additionally, CyberSecEval 2 expands on its predecessor by adding measures of an LLM’s propensity to allow for abuse of its code interpreter, offensive cybersecurity capabilities, and susceptibility to prompt injection attacks (learn more in our technical paper ). Finally, we’re introducing Code Shield which adds support for inference-time filtering of insecure code produced by LLMs. This offers mitigation of risks around insecure code suggestions, code interpreter abuse prevention, and secure command execution.

With the speed at which the generative AI space is moving, we believe an open approach is an important way to bring the ecosystem together and mitigate these potential harms. As part of that, we’re updating our Responsible Use Guide (RUG) that provides a comprehensive guide to responsible development with LLMs. As we outlined in the RUG, we recommend that all inputs and outputs be checked and filtered in accordance with content guidelines appropriate to the application. Additionally, many cloud service providers offer content moderation APIs and other tools for responsible deployment, and we encourage developers to also consider using these options.

Deploying Llama 3 at scale

Llama 3 will soon be available on all major platforms including cloud providers, model API providers, and much more. Llama 3 will be everywhere .

Our benchmarks show the tokenizer offers improved token efficiency, yielding up to 15% fewer tokens compared to Llama 2. Also, Group Query Attention (GQA) now has been added to Llama 3 8B as well. As a result, we observed that despite the model having 1B more parameters compared to Llama 2 7B, the improved tokenizer efficiency and GQA contribute to maintaining the inference efficiency on par with Llama 2 7B.

For examples of how to leverage all of these capabilities, check out Llama Recipes which contains all of our open source code that can be leveraged for everything from fine-tuning to deployment to model evaluation.

What’s next for Llama 3?

The Llama 3 8B and 70B models mark the beginning of what we plan to release for Llama 3. And there’s a lot more to come.

Our largest models are over 400B parameters and, while these models are still training, our team is excited about how they’re trending. Over the coming months, we’ll release multiple models with new capabilities including multimodality, the ability to converse in multiple languages, a much longer context window, and stronger overall capabilities. We will also publish a detailed research paper once we are done training Llama 3.

To give you a sneak preview for where these models are today as they continue training, we thought we could share some snapshots of how our largest LLM model is trending. Please note that this data is based on an early checkpoint of Llama 3 that is still training and these capabilities are not supported as part of the models released today.

We’re committed to the continued growth and development of an open AI ecosystem for releasing our models responsibly. We have long believed that openness leads to better, safer products, faster innovation, and a healthier overall market. This is good for Meta, and it is good for society. We’re taking a community-first approach with Llama 3, and starting today, these models are available on the leading cloud, hosting, and hardware platforms with many more to come.

Try Meta Llama 3 today

We’ve integrated our latest models into Meta AI, which we believe is the world’s leading AI assistant. It’s now built with Llama 3 technology and it’s available in more countries across our apps.

You can use Meta AI on Facebook, Instagram, WhatsApp, Messenger, and the web to get things done, learn, create, and connect with the things that matter to you. You can read more about the Meta AI experience here .

Visit the Llama 3 website to download the models and reference the Getting Started Guide for the latest list of all available platforms.

You’ll also soon be able to test multimodal Meta AI on our Ray-Ban Meta smart glasses.

As always, we look forward to seeing all the amazing products and experiences you will build with Meta Llama 3.

Our latest updates delivered to your inbox

Subscribe to our newsletter to keep up with Meta AI news, events, research breakthroughs, and more.

Join us in the pursuit of what’s possible with AI.

Product experiences

Foundational models

Latest news

Meta © 2024