#scalabledataextraction search results

Devvrit

11 h

Wish to build scaling laws for RL but not sure how to scale? Or what scales? Or would RL even scale predictably? We introduce: The Art of Scaling Reinforcement Learning Compute for LLMs

Devvrit_Khatri's tweet image. Wish to build scaling laws for RL but not sure how to scale? Or what scales? Or would RL even scale predictably?
We introduce: The Art of Scaling Reinforcement Learning Compute for LLMs

Google just released a python library for data extraction! LangExtract is a python library that extracts structured information from unstructured text documents with precise source grounding and interactive visualization. 100% Open Source

Sumanth_077's tweet image. Google just released a python library for data extraction!

LangExtract is a python library that extracts structured information from unstructured text documents with precise source grounding and interactive visualization.

100% Open Source

dataproducts

@dataproducts

11 h

Looking for ways to streamline your AI inferencing workflows and overwhelmed with unstructured data? Content-Aware IBM Storage Scale can help. stwb.co/eeeuaup

dataproducts's tweet image. Looking for ways to streamline your AI inferencing workflows and overwhelmed with unstructured data? Content-Aware IBM Storage Scale can help. stwb.co/eeeuaup

Unsloth AI

@UnslothAI

Oct 15

You can now train models up to 200B parameters locally on NVIDIA DGX Spark with Unsloth! 🦥 Fine-tune, RL & deploy OpenAI gpt-oss-120b via our free notebook in 68GB unified memory: colab.research.google.com/github/unsloth… Read our step-by-step guide in collab with NVIDIA docs.unsloth.ai/new/fine-tunin…

UnslothAI's tweet image. You can now train models up to 200B parameters locally on NVIDIA DGX Spark with Unsloth! 🦥

Fine-tune, RL &amp; deploy OpenAI gpt-oss-120b via our free notebook in 68GB unified memory: colab.research.google.com/github/unsloth…

Read our step-by-step guide in collab with NVIDIA docs.unsloth.ai/new/fine-tunin…

DECE Software

@SoftwareDECE

19 h

Too Big to Discover, No Time to Scan ⏱️📂 Your data may be petabytes in size — but your team doesn’t have weeks to wait. When time matters, sampling keeps discovery practical, fast, and effective. #DSPM #DataDiscovery #DataSecurity #GEODI

SoftwareDECE's tweet image. Too Big to Discover, No Time to Scan ⏱️📂

Your data may be petabytes in size — but your team doesn’t have weeks to wait.
When time matters, sampling keeps discovery practical, fast, and effective.

#DSPM #DataDiscovery #DataSecurity #GEODI

Nathan Lambert

@natolambert

14 h

The first fantastic paper on scaling RL with LLMs just dropped. I strongly recommend taking a look and will be sharing more thoughts on the blog soon. The Art of Scaling Reinforcement Learning Compute for LLMs Khatri & Madaan et al.

natolambert's tweet image. The first fantastic paper on scaling RL with LLMs just dropped. I strongly recommend taking a look and will be sharing more thoughts on the blog soon.

The Art of Scaling Reinforcement Learning Compute for LLMs
Khatri &amp; Madaan et al.

Om Singh

@om_kumar07

31 m

Just exported the @StanfordAIMI "CheXpert Dataset" (~450GB) to Local. can’t believe I pulled the entire dataset in just 8hrs on cellular data. This is an appreciation post for @reliancejio 👏 also thanks AzCopy for the seamless transfer @Azure

om_kumar07's tweet image. Just exported the @StanfordAIMI "CheXpert Dataset" (~450GB) to Local. can’t believe I pulled the entire dataset in just 8hrs on cellular data.
This is an appreciation post for @reliancejio 👏
also thanks AzCopy for the seamless transfer @Azure

Shubham Saboo

@Saboo_Shubham_

Aug 10

Google just released LangExtract Python library. It can extract structured data from unstructured docs with precise sources in just a few lines of code. 100% Opensource.

Saboo_Shubham_'s tweet image. Google just released LangExtract Python library.

It can extract structured data from unstructured docs with precise sources in just a few lines of code.

100% Opensource.

Joan Cabezas

@josancamon19

Oct 10

🧵 As AI labs race to scale RL, one question matters: when should you stop pre-training and start RL? We trained 5 Qwen models (0.6B→14B) with RL on GSM8K and found something wild: Small models see EMERGENCE-LIKE jumps. Large models see diminishing returns. The scaling law?…

josancamon19's tweet image. 🧵 As AI labs race to scale RL, one question matters: when should you stop pre-training and start RL?

We trained 5 Qwen models (0.6B→14B) with RL on GSM8K and found something wild:

Small models see EMERGENCE-LIKE jumps. Large models see diminishing returns.

The scaling law?…

elvis

@omarsar0

11 h

Banger paper from Meta and collaborators. This paper is one of the best deep dives yet on how reinforcement learning (RL) actually scales for LLMs. The team ran over 400,000 GPU hours of experiments to find a predictable scaling pattern and a stable recipe (ScaleRL) that…

omarsar0's tweet image. Banger paper from Meta and collaborators.

This paper is one of the best deep dives yet on how reinforcement learning (RL) actually scales for LLMs.

The team ran over 400,000 GPU hours of experiments to find a predictable scaling pattern and a stable recipe (ScaleRL) that…

Daily Dose of Data Science

@DailyDoseOfDS_

Sep 20

Google open-sourced LangExtract Python library! It uses LLMs to extract entities, attributes, and relations with exact source grounding from unstructured documents. Flexible LLM support (Gemini, OpenAI, Ollama) 100% open-source.

Firecrawl

@firecrawl_dev

Oct 15

Introducing Excel support You can now extract clean data from any Excel spreadsheet (.xlsx, xls) Try it out today in our API or playground 👇

firecrawl_dev's tweet image. Introducing Excel support

You can now extract clean data from any Excel spreadsheet (.xlsx, xls)

Try it out today in our API or playground 👇

Kirk Borne

@KirkDBorne

Oct 15

#DataEngineering with Scala and Spark — Build streaming & batch pipelines to process massive amounts of data: amzn.to/3wdmEhy v/ @PacktDataML 𝒦𝑒𝓎 𝐹𝑒𝒶𝓉𝓊𝓇𝑒𝓈: 🔴Transform data into a clean and trusted source of information for your organization using Scala…

KirkDBorne's tweet image. #DataEngineering with Scala and Spark — Build streaming &amp; batch pipelines to process massive amounts of data: amzn.to/3wdmEhy v/ @PacktDataML

𝒦𝑒𝓎 𝐹𝑒𝒶𝓉𝓊𝓇𝑒𝓈:

🔴Transform data into a clean and trusted source of information for your organization using Scala…

Vish

@vishjpat

5 h

Scaling RL with LLMs. ScaleRL is more scalable than prevalent RL methods. Algorithmic choices that seem to help generalization more Larger batch size , reducing truncations , longer generation lengths, and larger model scale .

vishjpat's tweet image. Scaling RL with LLMs.
ScaleRL is more scalable than prevalent RL methods.
Algorithmic choices that seem to help generalization more Larger batch size , reducing truncations , longer generation lengths, and larger model scale .

Marc Brooker

@MarcJBrooker

Feb 2, 2024

If we draw database rows as points, and add edges between rows that appear in the same transaction, the resulting graph is a great way to think about potential scalability. The more you can cut the graph up without crossing edges, the easier the workload is to scale.

MarcJBrooker's tweet image. If we draw database rows as points, and add edges between rows that appear in the same transaction, the resulting graph is a great way to think about potential scalability. The more you can cut the graph up without crossing edges, the easier the workload is to scale.

Devvrit

@Devvrit_Khatri

11 h

How do we understand the contribution of several design choices in an RL algorithm? Do they make the algorithm efficient? Or do they elevate the asymptotic performance? To study the scaling behavior of each design choice, we need to fit a predictable scaling curve - this provides…

Devvrit_Khatri's tweet image. How do we understand the contribution of several design choices in an RL algorithm? Do they make the algorithm efficient? Or do they elevate the asymptotic performance?
To study the scaling behavior of each design choice, we need to fit a predictable scaling curve - this provides…

ml-sanity bot

@arxivsanitybot

56 m

tinyurl.com/29bn5v35 Researchers unveil a groundbreaking framework for scaling reinforcement learning in large language models, introducing ScaleRL. After 400,000 GPU-hours of study, they provide a recipe for predictable RL training, paralleling pre-training's reliability.

arxivsanitybot's tweet image. tinyurl.com/29bn5v35 Researchers unveil a groundbreaking framework for scaling reinforcement learning in large language models, introducing ScaleRL. After 400,000 GPU-hours of study, they provide a recipe for predictable RL training, paralleling pre-training's reliability.

Devvrit

@Devvrit_Khatri

11 h

We provide (a) a framework to fit such scaling curves. Using this, we analyze several design choices, and combine the best ones to form our recipe (b) ScaleRL. We demonstrate its effectiveness by predictably scaling to 100k GPU-hours.

Devvrit_Khatri's tweet image. We provide (a) a framework to fit such scaling curves. Using this, we analyze several design choices, and combine the best ones to form our recipe (b) ScaleRL. We demonstrate its effectiveness by predictably scaling to 100k GPU-hours.

Yujia Qin

@TsingYoga

Jun 23, 2024

是时候把数据scale down了 LLaMA3告诉大家一个悲观的现实：模型架构不用动，把数据量从2T加到15T就可以暴力出奇迹。这一方面告诉大家基座模型长期来看就是大厂的机会；另一方面，考虑到scaling…

Data Scraping Services - Website Data Scraping

@_datascraping

18 h

Datascrapingservices.com is your partner in scalable data extraction. We provide the expertise needed to turn data challenges into opportunities for growth. Start your journey with us. Email [email protected]. #ScalableDataExtraction #Expertise #GrowthOpportunities

_datascraping's tweet image. Datascrapingservices.com is your partner in scalable data extraction. We provide the expertise needed to turn data challenges into opportunities for growth. Start your journey with us. Email info@datascrapingservices.com. #ScalableDataExtraction #Expertise #GrowthOpportunities

Data Scraping Services - Website Data Scraping

@_datascraping

Oct 10

Transform your business with Datascrapingservices.com. Our scalable data extraction services turn obstacles into growth opportunities and competitive advantages. Begin today at [email protected]. #TransformBusiness #ScalableDataExtraction #GrowthOpportunities

_datascraping's tweet image. Transform your business with Datascrapingservices.com. Our scalable data extraction services turn obstacles into growth opportunities and competitive advantages. Begin today at info@datascrapingservices.com. #TransformBusiness #ScalableDataExtraction #GrowthOpportunities

Data Scraping Services - Website Data Scraping

@_datascraping

Oct 8

Datascrapingservices.com provides scalable data extraction solutions to help you excel. Our expertise turns data challenges into growth opportunities. Connect with us today at [email protected]. #ScalableDataExtraction #Expertise #Excel #GrowthOpportunities