Home Technology peripherals AI Precision sniper Llama 3.1? Mistral AI open source Large 2, 123B is comparable to Llama 405B

Precision sniper Llama 3.1? Mistral AI open source Large 2, 123B is comparable to Llama 405B

Jul 25, 2024 pm 02:24 PM
industry

The

AI race is accelerating like never before, and following Meta’s launch yesterday of its new open source Llama 3.1 model, French AI startup Mistral AI has joined the fray.

Just now, Mistral AI announced the next generation of its flagship open source model: Mistral Large 2, which has 123 billion parameters and is on par with the latest cutting-edge models from OpenAI and Meta in terms of code generation, mathematics, reasoning, and more.

Following the release of Llama 3.1 405B, the release of Mistral Large 2 has suddenly made the track of open source large models lively, and the characteristic of this model is - "large enough".

Specifically, although the number of parameters of Mistral Large 2 is lower than Llama 3.1’s 405 billion, the performance of the two is close. And it’s comparable to GPT-4o and Anthropic’s Claude 3.5 Sonnet in multiple benchmarks.

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

In February this year, Mistral AI launched the original Large model, whose context window contained 32,000 tokens. The new version of the model is built on this basis and has a larger context window of 128,000 (approximately equivalent to a book) 300 page book) - Matches OpenAI's GPT-4o and GPT-4o mini and Meta's Llama 3.1.

Currently, Mistral Large 2 supports dozens of languages, including French, German, Spanish, Italian, Portuguese, Arabic, Hindi, Russian, Chinese, Japanese, and Korean, and more than 80 programming languages, including Python, Java, C, C++, JavaScript and Bash.

Mistral AI points out that new models will continue to push the boundaries of cost efficiency, speed and performance, while providing users with new capabilities, including advanced function calls and retrieval, to build high-performance AI applications.

However, it is worth noting that although Mistral Large 2 is open, it is limited to research and non-commercial use. It provides open weights, allowing third parties to fine-tune the model according to their needs. This agreement is an important limitation on the user's conditions of use. For commercial use that requires self-deployment of Mistral Large 2, a Mistral AI commercial license must be obtained in advance.

Performance

On multiple evaluation indicators, Mistral Large 2 has set new standards in performance and service cost. Especially on MMLU, the pre-trained version achieved an accuracy of 84.0%.

Code and Reasoning

Mistral AI trained Mistral Large 2 on a large portion of the code based on previous experience with Codestral 22B and Codestral Mamba.

Mistral Large 2 performs far better than the previous generation Mistral Large and is on par with top models like GPT-4o, Claude 3 Opus and Llama 3 405B.

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

Mistral AI has also put a lot of effort into enhancing the model's reasoning capabilities. One of the key points is to minimize the tendency of the model to produce "hallucinations" or produce information that seems reasonable but is actually incorrect or irrelevant. This is achieved by fine-tuning the model to be more careful and sharp in its responses, ensuring it delivers reliable and accurate output.

Additionally, Mistral Large 2 will admit when it cannot find a solution or does not have enough information to provide a confident answer. This pursuit of accuracy is reflected in improved model performance on math benchmarks, with the following graph demonstrating its enhanced reasoning and problem-solving capabilities:

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

                                                                                                                                      ‐ ‐ ‐ ‐ ‐ ‐ ​ ​ ​ ​  Models are all benchmarked through the same evaluation process). Performance accuracy on MultiPL-E (except paper, all models are benchmarked through the same evaluation process).

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

                          Performance accuracy on GSM8K (8-shot) and MATH (0-shot, no CoT) generation benchmarks (all models are benchmarked through the same evaluation process).

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405BInstructions to follow and align

ミストラル AI は、ミストラル ラージ 2 のコマンド追従機能と会話機能を大幅に向上させました。新しいミストラル ラージ 2 は、正確な指示に従い、複数ターンにわたる長い会話を処理することに特に優れています。以下は、MT-Bench、Wild Bench、および Arena Hard ベンチマーク テストでのパフォーマンスです:

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

ユニバーサル アライメント テストでのモデルのパフォーマンス (すべてのモデルは同じ評価 PIPELINE によってテストされます) 一部のベンチマークでは、より長い回答を生成すると評価が向上する傾向があります。ただし、多くの商用アプリケーションでは、簡潔なモデル生成により対話が高速化され、推論のコストが削減されるため、シンプルさが非常に重要です。

そのため、Mistral AIは、生成するコンテンツが可能な限り簡潔で簡潔であることを確認するために多大な努力を払っています。

次の図は、MT Bench ベンチマーク質問に対してさまざまなモデルによって生成された応答の平均長を示しています:

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

言語の多様性

今日の商用アプリケーションのシナリオの多くには、多言語ドキュメントの処理が含まれています。 Mistral Large 2 は、大量の多言語データ、特に英語、フランス語、ドイツ語、スペイン語、イタリア語、ポルトガル語、オランダ語、ロシア語、中国語、日本語、韓国語、アラビア語、ヒンディー語でトレーニングされました。

以下は、主に以前の Mistral Large、Llama 3.1 モデル、および Cohere の Command R+ と比較した、多言語 MMLU ベンチマークにおける Mistral Large 2 のパフォーマンス結果です。言語 MMLUパフォーマンス (ベースの事前トレーニング済みモデルで測定)

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405Bツールの使用法と関数呼び出し

Mistral Large 2 には、強化された関数呼び出しと取得スキルが装備されており、並列および逐次関数呼び出しを巧みに実行するようにトレーニングされています。は、複雑なビジネス アプリケーションの強力なエンジンとなる可能性があります。 精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B下の写真は、関数呼び出しに関して、Mistral Large 2と他の主流モデルの精度比較を示しています:

Mistral Large 2を試してください

ユーザーは、mistral-largeという名前のPlateformeでMistral Large 2を使用できます。 -2407、le Chat でテスト済み。バージョンは 24.07 (Mistral はすべてのモデルで YY.MM バージョン番号付けシステムを使用します)、API 名は mistral-large-2407 です。

指導モデルの重みは、HuggingFace でホストされて提供されます。

ウェイトリンク: https://huggingface.co/mistralai/Mistral-Large-Instruct-2407

Mistral AI は、Plateforme 上の製品を 2 つの一般的なモデルに統合しています:

Mistral Nemo精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B と Mistral Large、および 2 つのモデルプロフェッショナル モデル: Codestral と Embed。 Plateforme のように古いモデルが段階的に廃止される中、すべての Apache モデル (Mistral 7B、Mistral 8x7B および 8x22B、Codestral Mamba、Mathstral を含む) は引き続き、Mistral AI の SDK (mistral-inference および misstral-finetune Finetune) を使用してデプロイおよびデプロイできます。

本日より、Plateforme の微調整機能を拡張しました。現在は、Mistral Large、Mistral Nemo、Codestral に対応しています。 さらに、Mistral AI はクラウド サービス プロバイダーと協力しており、Mistral Large 2 は間もなくこれらのプラットフォームで利用可能になります。 Mistral AI は、Google Cloud Platform とのパートナーシップを拡大し、マネージド API を介して Mistral AI のモデルを Vertex AI に導入しました。それまでの間、Amazon Bedrock、Azure AI Studio、IBM watsonx.ai でも見つけることができます。

参考リンク:

https://mistral.ai/news/mistral-large-2407/

精准狙击Llama 3.1?Mistral AI开源Large 2,123B媲美Llama 405B

https://venturebeat.com/ai/mistral-ショック-with-new-open-model-mistral-large-2-aking-on-llama-3-1/

https://techcrunch.com/2024/07/24/mistral-releases- large-2-meta-openai-ai-models/

The above is the detailed content of Precision sniper Llama 3.1? Mistral AI open source Large 2, 123B is comparable to Llama 405B. For more information, please follow other related articles on the PHP Chinese website!

Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn

Hot AI Tools

Undresser.AI Undress

Undresser.AI Undress

AI-powered app for creating realistic nude photos

AI Clothes Remover

AI Clothes Remover

Online AI tool for removing clothes from photos.

Undress AI Tool

Undress AI Tool

Undress images for free

Clothoff.io

Clothoff.io

AI clothes remover

Video Face Swap

Video Face Swap

Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Roblox: Bubble Gum Simulator Infinity - How To Get And Use Royal Keys
4 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Nordhold: Fusion System, Explained
4 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
Mandragora: Whispers Of The Witch Tree - How To Unlock The Grappling Hook
3 weeks ago By 尊渡假赌尊渡假赌尊渡假赌

Hot Tools

Notepad++7.3.1

Notepad++7.3.1

Easy-to-use and free code editor

SublimeText3 Chinese version

SublimeText3 Chinese version

Chinese version, very easy to use

Zend Studio 13.0.1

Zend Studio 13.0.1

Powerful PHP integrated development environment

Dreamweaver CS6

Dreamweaver CS6

Visual web development tools

SublimeText3 Mac version

SublimeText3 Mac version

God-level code editing software (SublimeText3)

Hot Topics

Java Tutorial
1673
14
PHP Tutorial
1278
29
C# Tutorial
1257
24
DeepMind robot plays table tennis, and its forehand and backhand slip into the air, completely defeating human beginners DeepMind robot plays table tennis, and its forehand and backhand slip into the air, completely defeating human beginners Aug 09, 2024 pm 04:01 PM

But maybe he can’t defeat the old man in the park? The Paris Olympic Games are in full swing, and table tennis has attracted much attention. At the same time, robots have also made new breakthroughs in playing table tennis. Just now, DeepMind proposed the first learning robot agent that can reach the level of human amateur players in competitive table tennis. Paper address: https://arxiv.org/pdf/2408.03906 How good is the DeepMind robot at playing table tennis? Probably on par with human amateur players: both forehand and backhand: the opponent uses a variety of playing styles, and the robot can also withstand: receiving serves with different spins: However, the intensity of the game does not seem to be as intense as the old man in the park. For robots, table tennis

The first mechanical claw! Yuanluobao appeared at the 2024 World Robot Conference and released the first chess robot that can enter the home The first mechanical claw! Yuanluobao appeared at the 2024 World Robot Conference and released the first chess robot that can enter the home Aug 21, 2024 pm 07:33 PM

On August 21, the 2024 World Robot Conference was grandly held in Beijing. SenseTime's home robot brand "Yuanluobot SenseRobot" has unveiled its entire family of products, and recently released the Yuanluobot AI chess-playing robot - Chess Professional Edition (hereinafter referred to as "Yuanluobot SenseRobot"), becoming the world's first A chess robot for the home. As the third chess-playing robot product of Yuanluobo, the new Guoxiang robot has undergone a large number of special technical upgrades and innovations in AI and engineering machinery. For the first time, it has realized the ability to pick up three-dimensional chess pieces through mechanical claws on a home robot, and perform human-machine Functions such as chess playing, everyone playing chess, notation review, etc.

Claude has become lazy too! Netizen: Learn to give yourself a holiday Claude has become lazy too! Netizen: Learn to give yourself a holiday Sep 02, 2024 pm 01:56 PM

The start of school is about to begin, and it’s not just the students who are about to start the new semester who should take care of themselves, but also the large AI models. Some time ago, Reddit was filled with netizens complaining that Claude was getting lazy. "Its level has dropped a lot, it often pauses, and even the output becomes very short. In the first week of release, it could translate a full 4-page document at once, but now it can't even output half a page!" https:// www.reddit.com/r/ClaudeAI/comments/1by8rw8/something_just_feels_wrong_with_claude_in_the/ in a post titled "Totally disappointed with Claude", full of

At the World Robot Conference, this domestic robot carrying 'the hope of future elderly care' was surrounded At the World Robot Conference, this domestic robot carrying 'the hope of future elderly care' was surrounded Aug 22, 2024 pm 10:35 PM

At the World Robot Conference being held in Beijing, the display of humanoid robots has become the absolute focus of the scene. At the Stardust Intelligent booth, the AI ​​robot assistant S1 performed three major performances of dulcimer, martial arts, and calligraphy in one exhibition area, capable of both literary and martial arts. , attracted a large number of professional audiences and media. The elegant playing on the elastic strings allows the S1 to demonstrate fine operation and absolute control with speed, strength and precision. CCTV News conducted a special report on the imitation learning and intelligent control behind "Calligraphy". Company founder Lai Jie explained that behind the silky movements, the hardware side pursues the best force control and the most human-like body indicators (speed, load) etc.), but on the AI ​​side, the real movement data of people is collected, allowing the robot to become stronger when it encounters a strong situation and learn to evolve quickly. And agile

ACL 2024 Awards Announced: One of the Best Papers on Oracle Deciphering by HuaTech, GloVe Time Test Award ACL 2024 Awards Announced: One of the Best Papers on Oracle Deciphering by HuaTech, GloVe Time Test Award Aug 15, 2024 pm 04:37 PM

At this ACL conference, contributors have gained a lot. The six-day ACL2024 is being held in Bangkok, Thailand. ACL is the top international conference in the field of computational linguistics and natural language processing. It is organized by the International Association for Computational Linguistics and is held annually. ACL has always ranked first in academic influence in the field of NLP, and it is also a CCF-A recommended conference. This year's ACL conference is the 62nd and has received more than 400 cutting-edge works in the field of NLP. Yesterday afternoon, the conference announced the best paper and other awards. This time, there are 7 Best Paper Awards (two unpublished), 1 Best Theme Paper Award, and 35 Outstanding Paper Awards. The conference also awarded 3 Resource Paper Awards (ResourceAward) and Social Impact Award (

Li Feifei's team proposed ReKep to give robots spatial intelligence and integrate GPT-4o Li Feifei's team proposed ReKep to give robots spatial intelligence and integrate GPT-4o Sep 03, 2024 pm 05:18 PM

Deep integration of vision and robot learning. When two robot hands work together smoothly to fold clothes, pour tea, and pack shoes, coupled with the 1X humanoid robot NEO that has been making headlines recently, you may have a feeling: we seem to be entering the age of robots. In fact, these silky movements are the product of advanced robotic technology + exquisite frame design + multi-modal large models. We know that useful robots often require complex and exquisite interactions with the environment, and the environment can be represented as constraints in the spatial and temporal domains. For example, if you want a robot to pour tea, the robot first needs to grasp the handle of the teapot and keep it upright without spilling the tea, then move it smoothly until the mouth of the pot is aligned with the mouth of the cup, and then tilt the teapot at a certain angle. . this

Distributed Artificial Intelligence Conference DAI 2024 Call for Papers: Agent Day, Richard Sutton, the father of reinforcement learning, will attend! Yan Shuicheng, Sergey Levine and DeepMind scientists will give keynote speeches Distributed Artificial Intelligence Conference DAI 2024 Call for Papers: Agent Day, Richard Sutton, the father of reinforcement learning, will attend! Yan Shuicheng, Sergey Levine and DeepMind scientists will give keynote speeches Aug 22, 2024 pm 08:02 PM

Conference Introduction With the rapid development of science and technology, artificial intelligence has become an important force in promoting social progress. In this era, we are fortunate to witness and participate in the innovation and application of Distributed Artificial Intelligence (DAI). Distributed artificial intelligence is an important branch of the field of artificial intelligence, which has attracted more and more attention in recent years. Agents based on large language models (LLM) have suddenly emerged. By combining the powerful language understanding and generation capabilities of large models, they have shown great potential in natural language interaction, knowledge reasoning, task planning, etc. AIAgent is taking over the big language model and has become a hot topic in the current AI circle. Au

Hongmeng Smart Travel S9 and full-scenario new product launch conference, a number of blockbuster new products were released together Hongmeng Smart Travel S9 and full-scenario new product launch conference, a number of blockbuster new products were released together Aug 08, 2024 am 07:02 AM

This afternoon, Hongmeng Zhixing officially welcomed new brands and new cars. On August 6, Huawei held the Hongmeng Smart Xingxing S9 and Huawei full-scenario new product launch conference, bringing the panoramic smart flagship sedan Xiangjie S9, the new M7Pro and Huawei novaFlip, MatePad Pro 12.2 inches, the new MatePad Air, Huawei Bisheng With many new all-scenario smart products including the laser printer X1 series, FreeBuds6i, WATCHFIT3 and smart screen S5Pro, from smart travel, smart office to smart wear, Huawei continues to build a full-scenario smart ecosystem to bring consumers a smart experience of the Internet of Everything. Hongmeng Zhixing: In-depth empowerment to promote the upgrading of the smart car industry Huawei joins hands with Chinese automotive industry partners to provide

See all articles