• bitcoinBitcoin(BTC)$64,592.00-1.37%
  • ethereumEthereum(ETH)$1,806.86-1.48%
  • tetherTether(USDT)$1.000.00%
  • binancecoinBNB(BNB)$590.33-1.30%
  • usd-coinUSDC(USDC)$1.000.00%
  • rippleXRP(XRP)$0.96-1.22%
  • solanaSolana(SOL)$71.50-1.51%
  • tronTRON(TRX)$0.304550-1.30%
  • Figure HelocFigure Heloc(FIGR_HELOC)$1.060.00%
  • HyperliquidHyperliquid(HYPE)$66.80-2.05%
  • zcashZcash(ZEC)$893.33-3.28%
  • moneroMonero(XMR)$505.360.36%
  • dogecoinDogecoin(DOGE)$0.058825-0.90%
  • leo-tokenLEO Token(LEO)$9.211.34%
  • USDSUSDS(USDS)$1.000.00%
  • whitebitWhiteBIT Coin(WBT)$62.51-1.34%
  • chainlinkChainlink(LINK)$9.15-1.42%
  • cardanoCardano(ADA)$0.152483-2.04%
  • stellarStellar(XLM)$0.138300-0.85%
  • RainRain(RAIN)$0.00679984.84%
  • Ethena USDeEthena USDe(USDE)$1.00-0.01%
  • daiDai(DAI)$1.000.00%
  • nearNEAR Protocol(NEAR)$3.19-0.28%
  • bitcoin-cashBitcoin Cash(BCH)$205.64-2.07%
  • litecoinLitecoin(LTC)$44.47-2.40%
  • CantonCanton(CC)$0.0840264.22%
  • avalanche-2Avalanche(AVAX)$7.13-2.05%
  • uniswapUniswap(UNI)$5.14-0.91%
  • tether-goldTether Gold(XAUT)$3,857.690.94%
  • USD1USD1(USD1)$1.000.07%
  • suiSui(SUI)$0.72-1.60%
  • the-open-networkGram (prev. Toncoin)(GRAM)$1.04-2.12%
  • Ripple USDRipple USD(RLUSD)$1.000.01%
  • Global DollarGlobal Dollar(USDG)$1.000.01%
  • hedera-hashgraphHedera(HBAR)$0.058150-1.23%
  • paypal-usdPayPal USD(PYUSD)$1.00-0.01%
  • Ondo US Dollar YieldOndo US Dollar Yield(USDY)$1.150.04%
  • shiba-inuShiba Inu(SHIB)$0.000004-1.75%
  • quant-networkQuant(QNT)$154.281.14%
  • crypto-com-chainCronos(CRO)$0.042914-2.76%
  • BittensorBittensor(TAO)$188.98-2.24%
  • Circle USYCCircle USYC(USYC)$1.140.00%
  • Spiko Amundi Overnight Swap Fund (EUR)Spiko Amundi Overnight Swap Fund (EUR)(EURSAFO)$1.12-0.14%
  • MemeCoreMemeCore(M)$0.82-1.90%
  • okbOKB(OKB)$84.50-4.51%
  • pax-goldPAX Gold(PAXG)$3,859.790.89%
  • aaveAave(AAVE)$109.512.07%
  • Falcon USDFalcon USD(USDF)$1.000.06%
  • BlackRock USD Institutional Digital Liquidity FundBlackRock USD Institutional Digital Liquidity Fund(BUIDL)$1.000.00%
  • United StablesUnited Stables(U)$1.000.00%
TradePoint.io
  • Main
  • AI & Technology
  • Stock Charts
  • Market & News
  • Business
  • Finance Tips
  • Trade Tube
  • Blog
  • Shop
No Result
View All Result
TradePoint.io
No Result
View All Result

NuMind Releases Three SOTA NER Models that Outperform Similar-Sized Foundation Models in the Few-shot Regime and Competing with Much Larger LLMs

May 17, 2024
in AI & Technology
Reading Time: 5 mins read
A A
NuMind Releases Three SOTA NER Models that Outperform Similar-Sized Foundation Models in the Few-shot Regime and Competing with Much Larger LLMs
ShareShareShareShareShare




Named Entity Recognition (NER) is vital in natural language processing, with applications spanning medical coding, financial analysis, and legal document parsing. Custom models are typically created using transformer encoders pre-trained on self-supervised tasks like masked language modeling (MLM). However, recent years have seen the rise of large language models (LLMs) like GPT-3 and GPT-4, which can tackle NER tasks through well-crafted prompts but pose challenges due to high inference costs and potential privacy concerns.

NuMind team introduces an approach that suggests utilizing LLMs to minimize human annotations for custom model creation. Rather than employing an LLM to annotate a single-domain dataset for a specific NER task, the idea involves using the LLM to annotate a diverse, multi-domain dataset covering various NER problems. Subsequently, a smaller foundation model like BERT is further pre-trained on this annotated dataset. This pre-trained model can then be fine-tuned for any downstream NER task.

YOU MAY ALSO LIKE

Moth’s Tessa Encodes and Recovers 1,568 Values on a 120-Qubit Chip – Unite.AI

Flagship Pioneering Launches London Lab to Build AI-Native Companies – Unite.AI

The team has introduced its three NER models, which are the following:

  1. NuNER Zero: A zero-shot NER model adopts the GLiNER (Generalist Model for Named Entity Recognition using Bidirectional Transformer) architecture and requires input as a concatenation of entity types and text. Unlike GLiNER, NuNER Zero functions as a token classifier, enabling the detection of arbitrarily long entities. Trained on the NuNER v2.0 dataset, which merges subsets of Pile and C4 annotated via LLMs using NuNER’s procedure, NuNER Zero emerges as the leading compact zero-shot NER model, boasting a +3.1% token-level F1-Score improvement over GLiNER-large-v2.1 on GLiNER’s benchmark.
  1. NuNER Zero 4k: NuNER Zero 4k is the long-context (4k tokens) version of NuNER Zero. It is generally less performant than NuNER Zero but can outperform NuNER Zero on applications where context size matters.
  1. NuNER Zero-span: NuNER Zero-span is the span-prediction version of NuNER Zero, which shows slightly better performance than NuNER Zero but cannot detect entities larger than 12 tokens.

The key features of these three models are:

  • NuNER Zero: Originated from NuNER, convenient for moderate token size.
  • NuNER Zero 4K: A variation of NuNER performs better in scenarios where context size matters.
  • NuNER Zero-span: The span-prediction version of NuNER Zero is not convenient for entities larger than 12 tokens.

In conclusion, NER is crucial in natural language processing, yet creating custom models typically relies on transformer encoders trained via MLM. However, the rise of LLMs like GPT-3 and GPT-4 poses challenges due to high inference costs. The NuMind team proposes an approach utilizing LLMs to reduce human annotations by annotating a multi-domain dataset. They introduce three NER models: NuNER Zero, a compact zero-shot model; NuNER Zero 4k, emphasizing longer context; and NuNER Zero-span, prioritizing span prediction with slight performance enhancements but limited to entities under 12 tokens.


Sources

  1. https://huggingface.co/numind/NuNER_Zero-4k
  2. https://huggingface.co/numind/NuNER_Zero
  3. https://huggingface.co/numind/NuNER_Zero-span
  4. https://arxiv.org/pdf/2402.15343
  5. https://www.linkedin.com/posts/tomaarsen_numind-yc-s22-has-just-released-3-new-state-of-the-art-activity-7195863382783049729-kqko/?utm_source=share&utm_medium=member_ios


Sana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions.


🐝 Join the Fastest Growing AI Research Newsletter Read by Researchers from Google + NVIDIA + Meta + Stanford + MIT + Microsoft and many others…






Previous articleAgentClinic: Simulating Clinical Environments for Assessing Language Models in Healthcare
Next articleTop AI Tools for Real Estate Agents


Credit: Source link

ShareTweetSendSharePin

Related Posts

Moth’s Tessa Encodes and Recovers 1,568 Values on a 120-Qubit Chip – Unite.AI
AI & Technology

Moth’s Tessa Encodes and Recovers 1,568 Values on a 120-Qubit Chip – Unite.AI

November 16, 2026
Flagship Pioneering Launches London Lab to Build AI-Native Companies – Unite.AI
AI & Technology

Flagship Pioneering Launches London Lab to Build AI-Native Companies – Unite.AI

November 16, 2026
Apple’s New Research App Is Designed To Make AI Health Tools More User-Friendly
AI & Technology

Apple’s New Research App Is Designed To Make AI Health Tools More User-Friendly

November 16, 2026
APA Issues Workplace AI Guidance on Trust, Oversight and Job Design – Unite.AI
AI & Technology

APA Issues Workplace AI Guidance on Trust, Oversight and Job Design – Unite.AI

November 16, 2026
Next Post
French police killed suspect planning to burn synagogue in Rouen

French police killed suspect planning to burn synagogue in Rouen

Leave a Reply Cancel reply

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

Search

No Result
View All Result
Kornacki kicks off all-night election results stream

Kornacki kicks off all-night election results stream

November 11, 2026
How to Write a Check So No One Can Change It

How to Write a Check So No One Can Change It

November 14, 2026
Morning News NOW Full Episode – Nov. 3

Morning News NOW Full Episode – Nov. 3

November 11, 2026

About

Learn more

Our Services

Legal

Privacy Policy

Terms of Use

Bloggers

Learn more

Article Links

Contact

Advertise

Ask us anything

©2020- TradePoint.io - All rights reserved!

Tradepoint.io, being just a publishing and technology platform, is not a registered broker-dealer or investment adviser. So we do not provide investment advice. Rather, brokerage services are provided to clients of Tradepoint.io by independent SEC-registered broker-dealers and members of FINRA/SIPC. Every form of investing carries some risk and past performance is not a guarantee of future results. “Tradepoint.io“, “Instant Investing” and “My Trading Tools” are registered trademarks of Apperbuild, LLC.

This website is operated by Apperbuild, LLC. We have no link to any brokerage firm and we do not provide investment advice. Every information and resource we provide is solely for the education of our readers. © 2020 Apperbuild, LLC. All rights reserved.

No Result
View All Result
  • Main
  • AI & Technology
  • Stock Charts
  • Market & News
  • Business
  • Finance Tips
  • Trade Tube
  • Blog
  • Shop

© 2023 - TradePoint.io - All Rights Reserved!