NTT 上席特別研究員 西田氏のご講演「NTT版大規模言語モデル「tsuzumi」の取組について」でLLM分野の今後の展望についてキャッチアップしましょう! #FullyConnectedTokyo イベント参加登録はこちらから→ fullyconnected.jp Come learn from Kyosuke Nishida, Senior Distinguished Researcher at NTT, and other builders of the Generative AI industry at Fully Connected on October 10th in Tokyo. Register here → fullyconnected.jp
Weights & Biases
Software Development
San Francisco, California 73,460 followers
The AI developer platform.
About us
Weights & Biases: the AI developer platform. Build better models faster, fine-tune LLMs, develop GenAI applications with confidence, all in one system of record developers are excited to use. W&B Models is the MLOps solution used by foundation model builders and enterprises who are training, fine-tuning, and deploying models into production. W&B Weave is the LLMOps solution for software developers who want a lightweight but powerful toolset to help them track and evaluate LLM applications. Weights & Biases is trusted by over a 1,000 companies to productionize AI at scale including teams at OpenAI, Meta, NVIDIA, Cohere, Toyota, Square, Salesforce, and Microsoft. Sign up for a 30-day free trial today at http://wandb.me/trial.
- Website
-
https://wandb.ai/site
External link for Weights & Biases
- Industry
- Software Development
- Company size
- 201-500 employees
- Headquarters
- San Francisco, California
- Type
- Privately Held
- Founded
- 2017
- Specialties
- deep learning, developer tools, machine learning, MLOps, GenAI, LLMOps, large language models, and llms
Products
Weights & Biases
Machine Learning Software
Weights & Biases helps AI developers build better models faster. Quickly track experiments, version and iterate on datasets, evaluate model performance, reproduce models, and manage your ML workflows end-to-end.
Locations
-
Primary
400 Alabama St
San Francisco, California 94110, US
Employees at Weights & Biases
Updates
-
Weights & Biases reposted this
⚡ AI Hacker Cup Lightning Competition Today we're kicking off a 7-day competition to solve all 5 of the 2023 practice Hacker Cup challenges with Mistral AI models. Our current baseline is 2/5 with the starter RAG agent (with reflection) We're also providing Mistral AI api access Details 👉 https://lnkd.in/eGMEkcnp Prizes We're giving out 3 x Meta Ray Bans - one for the first individual to hit 3/5, 4/5 & 5/5 problems solved. The Challenge Meta has been running the Hacker Cup since 2011 - its programming challenge where a set of problems must be solved in a fixed amount of time. This year Mark Saroufim (PyTorch) and Weiwei Yang (Microsoft Research) are taking an AI track to NeurIPS. Its unclear if AI systems will be able to beat humans at this task. We're running lightning competition ahead of the official 2024 practice round challenges being posted to get people warmed up for the main event. Starter Kit Last week Ramanathan Parameshwaran described the RAG code generation agent in the starter kit - it includes: 1️⃣ attempting a zero shot attempt 2️⃣ if not, then using the CodeContests dataset (from Google DeepMind) as few shot examples 3️⃣ if not, then reflecting on the mistakes and trying again 📺 You can see his full breakdown here: https://lnkd.in/euA_GAjH Evaluations We're using Weave Evaluations from Weights & Biases to run evals as well as help verify submissions. We're going on a certain amount of trust here but there are some straightforward ways to report 5/5 in this pipeline and Weave (plus code submission) will help us verify that submissions are legit. Ok enough, time to hack, come chat in the official NeurIPS AI Hacker Cup discord: https://lnkd.in/eR84CXzS And try tweak the prompts in the starter kit to see how far that can take you. (huge props to Ramanathan Parameshwaran and Thomas Capelle for getting this starter kit off the ground 🙏 )
-
Join the GenAI Salon this week in San Francisco with the CEO of Replicate, Ben Fishman. Learn how Replicate grew to millions of users and insights from AI projects in production. Space is limited—register now: https://lnkd.in/g38y8B6M
-
How do you stay ahead in GenAI? Join Weights & Biases’ Nicolas Remerscheid at PyData Munich's MLOps Meetup, where he’ll share essential LLMOps insights for faster, more confident development with strategic takeaways and real-world examples. Register now: https://lnkd.in/gkCrC_99
-
Get hands-on experience with improving data ingestion, retrieval, reranking, and more using the Cohere API for advanced RAG with the soon-to-be-released RAG++ course: https://lnkd.in/ggWwT-ht Special thanks to Cohere for providing credits to course participants.
-
深層学習モデル「トランスフォーマー」の論文の著者の1人であり、Sakana AI AI Co-Founder and CTO Llion Jones氏のJapanを拠点とした次の一手に注目しましょう!#FullyConnectedTokyo イベント参加登録はこちらから→ fullyconnected.jp Come learn from Llion Jones, Co-Founder and CTO at Sakana AI, and other builders of the Generative AI industry at Fully Connected on October 10th in Tokyo Register here → fullyconnected.jp #FullyConnectedTokyo
-
Optimize your RAG pipeline with task-specific data ingestion & chunking techniques. Learn how to use metadata extraction, parent-document retrieval, & best practices for ensuring high-quality retrieval with our soon to be launched RAG ++ course: https://lnkd.in/ggWwT-ht
-
Weights & Biases reposted this
Join me in a couple of weeks (9/18) to discuss how Weights & Biases can help your company establish a System of Record for your Machine Learning workloads! https://lnkd.in/geUqADXZ
Building a system of record for ML - Webinar
https://site.wandb.ai
-
Join us at the in-person W&B Hackathon to build and improve LLM Judges. Whether you’re refining an existing model or creating a new annotation UI, this event is for AI Engineers who are ready to push the boundaries. Cash prizes and LLM API credits available. Register here: https://lnkd.in/g8VXptg7