Tootfinder

Well that is very interesting! Introducing Mistral-Large on Azure in partnership with Mistral AI | Microsoft Azure Blog
https://azure.microsoft.com/en-us/blog/microsof…

Introducing Mistral-Large on Azure in partnership with Mistral AI | Microsoft Azure Blog
Microsoft is announcing a multi-year partnership with Mistral AI, bridging the gap between pioneering research and real-world solutions.

Multi-purpose InSTRument for Astronomy at Low-resolution: MISTRAL@OHP
J. Schmitt, C. Adami, M. Dennefeld, F. Agneray, S. Basa, J. C. Brunel, V. Buat, D. Burgarella, C. Carvalho, G. Castagnoli, N. Grosso, F. Huppert, C. Moreau, F. Moreau, L. Moreau, E. Muslimov, S. Pascal, S. Perruchot, D. Russeil, J. L. Beuzit, F. Dolon, M. Ferrari, B. Hamelin, A. LevanSuu, K. Aravind, D. Gotz, E. Jehin, E. LeFloc'h, J. Palmerio, A. Saccardi, B. Schneider, F. Sch\"ussler, D. Turpin, S. D. Verg…

Multi-purpose InSTRument for Astronomy at Low-resolution: MISTRAL@OHP
MISTRAL is the new Faint Object Spectroscopic Camera mounted at the folded Cassegrain focus of the 1.93m telescope of Haute-Provence Observatory. We describe the design and components of the instrument and give some details about its operation. We emphasise in particular the various observing modes and the performances of the detector. A short description is also given about the working environment. Various types of objects, including stars, nebulae, comets, novae, galaxies have been observed d…

KI-Update kompakt: Mistral, Nvidia RTX, KI-Konzepthandy, Umweltkosten von KI
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
https://www.

Patronus AI releases CopyrightCatcher and says GPT-4 produced copyrighted content on 44% of prompts, Mistral's Mixtral on 22%, Llama 2 10%, and Claude 2.1 8% (Hayden Field/CNBC)
https://www.cnbc.com/2024/03/06/gpt-4-rese

#Paris I like his responses, he is more direct to answer, not loong sentences like #gpt

Simple Techniques for Enhancing Sentence Embeddings in Generative Language Models
Bowen Zhang, Kehua Chang, Chunping Li
https://arxiv.org/abs/2404.03921 ht…

Simple Techniques for Enhancing Sentence Embeddings in Generative Language Models
Sentence Embedding stands as a fundamental task within the realm of Natural Language Processing, finding extensive application in search engines, expert systems, and question-and-answer platforms. With the continuous evolution of large language models such as LLaMA and Mistral, research on sentence embedding has recently achieved notable breakthroughs. However, these advancements mainly pertain to fine-tuning scenarios, leaving explorations into computationally efficient direct inference method…

Another Chatbot, this time using Mistral, #AI #Mistral

Microsoft says its investment in Mistral AI amounts to €15M; EU regulators plan to analyze the investment, which could eventually lead to a formal investigation (Samuel Stolton/Bloomberg)
https://www.bloomberg.com/news/articles/20

Microsoft’s Mistral AI Investment to Be Examined by EU Watchdog
Microsoft Corp.’s Mistral AI investment is set to be analyzed by the European Union’s competition watchdog at the same time that its deep ties to OpenAI Inc come under regulatory scrutiny.

https://techcrunch.com/2024/02/26/mistral-ai-releases-new-model-to-rival-gpt-4-and-its-own-chat-assistant/?eicker.news
In December I tried the small model. Really good, also in ressource consumption.…

Mistral AI releases new model to rival GPT-4 and its own chat assistant | TechCrunch
Mistral AI is launching a new flagship large language model called Mistral Large. It is designed to rival other top-tier models like GPT-4.

#Paris I like his responses, he is more direct to answer, not loong sentences like #gpt

SaulLM-7B: A pioneering Large Language Model for Law
Pierre Colombo, Telmo Pessoa Pires, Malik Boudiaf, Dominic Culver, Rui Melo, Caio Corro, Andre F. T. Martins, Fabrizio Esposito, Vera L\'ucia Raposo, Sofia Morgado, Michael Desa
https://arxiv.org/abs/2403.03883

SaulLM-7B: A pioneering Large Language Model for Law
In this paper, we introduce SaulLM-7B, a large language model (LLM) tailored for the legal domain. With 7 billion parameters, SaulLM-7B is the first LLM designed explicitly for legal text comprehension and generation. Leveraging the Mistral 7B architecture as its foundation, SaulLM-7B is trained on an English legal corpus of over 30 billion tokens. SaulLM-7B exhibits state-of-the-art proficiency in understanding and processing legal documents. Additionally, we present a novel instructional fine…

KI-Update kompakt: Gemini 1.5, Mistral Next, Altmans Chip-Pläne, Copilot
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
https://www.

Microsoft partners with Mistral as regulators probe its OpenAI alliance; Microsoft takes a minor stake in Mistral and help bring Mistral's AI models to market (Madhumita Murgia/Financial Times)
https://t.co/9Jk7fp65JN

This https://arxiv.org/abs/2403.03883 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

SaulLM-7B: A pioneering Large Language Model for Law
In this paper, we introduce SaulLM-7B, a large language model (LLM) tailored for the legal domain. With 7 billion parameters, SaulLM-7B is the first LLM designed explicitly for legal text comprehension and generation. Leveraging the Mistral 7B architecture as its foundation, SaulLM-7B is trained on an English legal corpus of over 30 billion tokens. SaulLM-7B exhibits state-of-the-art proficiency in understanding and processing legal documents. Additionally, we present a novel instructional fine…

Gemma gets very bad reviews and Mistral ( Mixtral 8x7B) seems much better. What is Google doing ?
#Google #Gemma #Mistral

Mixtral of experts
A high quality Sparse Mixture-of-Experts.

#AI at it‘s best …

@… yeah I know, thanks. I already switched to gpt4, I only ment: I still have to test the Mistral model :)

This is fine… it's fine! https://www.ft.com/content/cd6eb51a-3276-450f-87fd-97e8410db9eb

Gemma gets very bad reviews and Mistral ( Mixtral 8x7B) seems much better. What is Google doing ?
#Google #Gemma #Mistral

Mixtral of experts
A high quality Sparse Mixture-of-Experts.

Dienstag: Microsoft mit neuem Partner, Mond-Lander in Seitenlage schickt Bilder
Microsoft investiert in Mistral Mond-Lander funkt Bilder KI als Programmierer-Ersatz LinkedIn nutzt sensible Daten New Outlook statt Mail und Kalender

Finally a good model for many languages.
Question to mixtral-8x7b-v01-q
Mit welchen 25 Sprachen wurde das Mixtral 8x7b instruct trainiert? Welche dieser Sprachen eignen sich um mit dem Modell zu kommunizieren?
#Mistral #mixtral-8x7b-v01-q

Here we go into the future!
I spent the day with the new Mistral Large API and I asked it to create two songs for my podcast.
One in Syncopated Trap style and the other in Gypsy Jazz.
Here's the Syncopated Trap version.
#AI #AImusic

Elon Musk might be suing OpenAI because he is a troll and has a grudge but his lawsuit puts a spotlight on the empty humanitarian posturing by AI businesses (Parmy Olson/Bloomberg)
https://www.bloomberg.com/opinion/articles

Elon Musk Is Right to Call Out Sam Altman’s OpenAI Hypocrisy
Too many AI startups have seen their noble goals imperiled by Big Tech’s gravitational pull.

Workshopfrage letzte Woche: "Wie kann ich selber ein offenes LLM z.B auf meinem Laptop austesten, wenn ich keine GPU habe?"
⏩ Ollama.ai kann Modelle auch nur mit einer CPU ausführen. Das genügt zum Testen (Output wird langsam generiert), man braucht nur Festplattenplatz.
Wer ein LLM auf Deutsch testen möchte, kann dann einfach > ollama run mixtral ausführen. Über die Modelfile-Konfigurationsdatei lässt sich via System Prompt einstellen, wie das Modell antworten soll…

ollama/docs/modelfile.md at main · ollama/ollama
Get up and running with Llama 2, Mistral, Gemma, and other large language models. - ollama/ollama

#AI at it‘s best …

Source: Mistral is in talks with investors about raising several hundred million dollars at a $5B valuation, after raising $415M at a $2B valuation in Dec. 2023 (Kate Clark/The Information)
https://www.theinformation.com/articles/mi

Mistral, an OpenAI Rival in Europe, in Talks to Raise Capital at a $5 Billion Valuation
Mistral, a Paris-based startup developing artificial intelligence that is open source, has been speaking to investors about raising several hundred million dollars at a valuation of $5 billion, according to a person with direct knowledge. The company, which has just recently begun to generate ...

Introducing Super RAGs in Mistral 8x7B-v1
Ayush Thakur, Raghav Gupta
https://arxiv.org/abs/2404.08940 https://arxiv.org/pdf/2404.0894…

Introducing Super RAGs in Mistral 8x7B-v1
The relentless pursuit of enhancing Large Language Models (LLMs) has led to the advent of Super Retrieval-Augmented Generation (Super RAGs), a novel approach designed to elevate the performance of LLMs by integrating external knowledge sources with minimal structural modifications. This paper presents the integration of Super RAGs into the Mistral 8x7B v1, a state-of-the-art LLM, and examines the resultant improvements in accuracy, speed, and user satisfaction. Our methodology uses a fine-tuned…

Here we go into the future: Part 2!
As you know from my previous post, I spent the day with the new Mistral Large API and I asked it to create two songs for my podcast.
One in Syncopated Trap style and the other in Gypsy Jazz.
Here's the Gypsy Jazz version.
#AI #AImusic

I love how "you can't criticize Mistral for hollowing out the #AIAct while selling itself to Microsoft at the same time because without Hyperscaler infrastructure #AI doesn't work" is a thing people say now.
Like use your imagination to go beyond "in the future you can choose…

Nick Clegg, who could have insisted on PR instead of FPTP, but instead enabled Cameron to be an awful PM who caused Brexit, is now talking nonsence about AI
https://apnews.com/article/meta-ai-assistant-llama3-large-lan…

Facebook parent Meta Platforms unveils new set of artificial intelligence systems
Generative AI is advancing so quickly that the latest chatbots available today could be out of date tomorrow. Google, Meta Platforms and OpenAI, along with startups such as Anthropic, Cohere and France’s Mistral, have been churning out new AI language models and hoping to persuade customers they’ve got the smartest, handiest or most efficient chatbots. Meta is the latest to up its game, unveiling new models Thursday that will be among the most visible: they’re already getting baked into F…

New releases for the major AI models.
https://www.theguardian.com/technology/2024/apr/10/ai-race-heats-up-as-openai-google-and-mistral-release-new-models

How I Reduced Our #LLM #Costs by Over 85%
https://jellypod.ai/blog/how-I-reduced-llm-costs

A profile of Mistral AI CEO Arthur Mensch, who says, as an atheist, he is uncomfortable with Silicon Valley's "AGI rhetoric" and "religious" fascination with AI (New York Times)
…

Europe’s A.I. ‘Champion’ Sets Sights on Tech Giants in U.S.
Mistral, a French start-up considered a promising challenger to OpenAI and Google, is getting support from European leaders who want to protect the region’s culture and politics.

Sentinels of the Stream: Unleashing Large Language Models for Dynamic Packet Classification in Software Defined Networks -- Position Paper
Shariq Murtuza
https://arxiv.org/abs/2402.07950

Sentinels of the Stream: Unleashing Large Language Models for Dynamic Packet Classification in Software Defined Networks -- Position Paper
With the release of OpenAI's ChatGPT, the field of large language models (LLM) saw an increase of academic interest in GPT based chat assistants. In the next few months multiple accesible large language models were released that included Meta's LLama models and Mistral AI's Mistral and Mixtral MoE models. These models are available openly for a wide array of purposes with a wide spectrum of licenses. These LLMs have found their use in a different number of fields like code development, SQL gene…

We review the high end Wolf Coffee machine. We examine why Tyler Perry is right to embrace AI entertainment. We compare API friendliness for Anthropic, Gemini, Mistral and Perplexity!
Here's the question we asked all our API Bots in our test:
"Where do you start to write a story? Explain your process. As well, tell us why the name 'Elara' is so often used by LLMs in creative storytelling as a name."

KI-Unternehmen zum AI Act: Lobbyismus oder unangebrachtes Eingreifen?
Aleph Alpha und Mistral haben ihre Nähe zu den Regierungen genutzt, um den AI Act mitzugestalten. Lobby Control sieht zu viel Eingreifen.

Trying LM Studio 0.2.16 for AMD ROCm (Technology Preview Release) Mistral for the first time. Got myself my own local AI!
I don't know what to do now 😄
#AI #LMStudio #AMDROCm

A profile of Mistral AI CEO Arthur Mensch, who says, as an atheist, he is uncomfortable with Silicon Valley's "AGI rhetoric" and "religious" fascination with AI (New York Times)
…

Europe’s A.I. ‘Champion’ Sets Sights on Tech Giants in U.S.
Mistral, a French start-up considered a promising challenger to OpenAI and Google, is getting support from European leaders who want to protect the region’s culture and politics.

Meta claims both Llama 3 models beat similarly sized models like Gemini, Mistral, and Claude 3 in certain benchmarks; humans marked Llama 3 higher than GPT-3.5 (Emilia David/The Verge)
https://www.theverge.com/2024/4/18/24134103/llama-…

Meta says Llama 3 beats most other models, including Gemini
Meta says Llama 3 outperformed models like Google’s Gemma and Gemini and OpenAI’s GPT-3.5 in several benchmark tests.

Verif.ai: Towards an Open-Source Scientific Generative Question-Answering System with Referenced and Verifiable Answers
Milo\v{s} Ko\v{s}prdi\'c, Adela Ljaji\'c, Bojana Ba\v{s}aragin, Darija Medvecki, Nikola Milo\v{s}evi\'c
https://arxiv.org/abs/2402.18589

Verif.ai: Towards an Open-Source Scientific Generative Question-Answering System with Referenced and Verifiable Answers
In this paper, we present the current progress of the project Verif.ai, an open-source scientific generative question-answering system with referenced and verified answers. The components of the system are (1) an information retrieval system combining semantic and lexical search techniques over scientific papers (PubMed), (2) a fine-tuned generative model (Mistral 7B) taking top answers and generating answers with references to the papers from which the claim was derived, and (3) a verification…

@… yes, Claude is amazing as well, but phind is able to search the web (like Perplexity.ai) - that's a real benefit. When it comes to coding questings, phind - my opinion - shows better results than gpt4, Mistral and Perplexity.ai

Sources: Mistral plans to raise €500M at an over €5B valuation from multiple VCs and wealth funds, after raising ~€400M at a €2B valuation in December 2023 (Financial Times)
https://t.co/aldaMDD3Jz

This https://arxiv.org/abs/2402.12147 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Surprising Efficacy of Fine-Tuned Transformers for Fact-Checking over Larger Language Models
In this paper, we explore the challenges associated with establishing an end-to-end fact-checking pipeline in a real-world context, covering over 90 languages. Our real-world experimental benchmarks demonstrate that fine-tuning Transformer models specifically for fact-checking tasks, such as claim detection and veracity prediction, provide superior performance over large language models (LLMs) like GPT-4, GPT-3.5-Turbo, and Mistral-7b. However, we illustrate that LLMs excel in generative tasks …

Do Large Language Models Learn Human-Like Strategic Preferences?
Jesse Roberts, Kyle Moore, Doug Fisher
https://arxiv.org/abs/2404.08710 https://

Do Large Language Models Learn Human-Like Strategic Preferences?
We evaluate whether LLMs learn to make human-like preference judgements in strategic scenarios as compared with known empirical results. We show that Solar and Mistral exhibit stable value-based preference consistent with human in the prisoner's dilemma, including stake-size effect, and traveler's dilemma, including penalty-size effect. We establish a relationship between model size, value based preference, and superficiality. Finally, we find that models that tend to be less brittle were train…

Microsoft strikes a deal with Mistral.
#AI #GenerativeAI #LLM #Microsoft

Microsoft partners with Mistral in second AI deal beyond OpenAI
Microsoft is partnering with French AI startup Mistral to bring more AI models to its Azure platform. It’s the second major AI investment for Microsoft.

Mistral AI launches Mixtral 8x22B, its latest sparse mixture-of-experts model, after releasing Mixtral 8x7B in December 2023 (Shubham Sharma/VentureBeat)
https://venturebeat.com/ai/mistral-ai-drops-new-mixture-of-experts-model-with-a-t…

This https://arxiv.org/abs/2402.10693 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

Exploring Precision and Recall to assess the quality and diversity of LLMs
This paper introduces a novel evaluation framework for Large Language Models (LLMs) such as Llama-2 and Mistral, focusing on the adaptation of Precision and Recall metrics from image generation to text generation. This approach allows for a nuanced assessment of the quality and diversity of generated text without the need for aligned corpora. By conducting a comprehensive evaluation of state-of-the-art language models, the study reveals significant insights into their performance on open-ended …

https://www.youtube.com/watch?v=3iZ7jwbAwF8
Embedding Layer of mistral-7b and gemma-7b
Chris Hay plays with the models. Good practical insights with hands on using python.

Interessant, bij #Perplexity in een labs omgeving de wat kleinere versies van de bekende modellen online beschikbaar, Claude, Gemma, Mistral etc.
https://labs.perplexity.ai/

This https://arxiv.org/abs/2402.13116 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

A Survey on Knowledge Distillation of Large Language Models
This survey presents an in-depth exploration of knowledge distillation (KD) techniques within the realm of Large Language Models (LLMs), spotlighting the pivotal role of KD in transferring sophisticated capabilities from proprietary giants such as GPT-4 to accessible, open-source models like LLaMA and Mistral. Amidst the evolving AI landscape, this work elucidates the critical disparities between proprietary and open-source LLMs, demonstrating how KD serves as an essential conduit for imbuing t…

Me, when I am talking with my personal LLM.
#AI #LLM #Mistral #StarTrek

Mistral.rs: A Lightning-Fast #LLM #Inference Platform with Device Support, Quantization, and #OpenAI #API

The GPT-4 barrier has finally been broken, with Gemini 1.5, Mistral Large, Claude 3 Opus, and Inflection-2.5 benchmarking near or even above OpenAI's model (Simon Willison's Weblog)
https://simonwillison.net/2024/Mar/8/gpt-4-barrier/

This https://arxiv.org/abs/2402.13116 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCL_…

A Survey on Knowledge Distillation of Large Language Models
In the era of Large Language Models (LLMs), Knowledge Distillation (KD) emerges as a pivotal methodology for transferring advanced capabilities from leading proprietary LLMs, such as GPT-4, to their open-source counterparts like LLaMA and Mistral. Additionally, as open-source LLMs flourish, KD plays a crucial role in both compressing these models, and facilitating their self-improvement by employing themselves as teachers. This paper presents a comprehensive survey of KD's role within the realm…