Home Blog Page 1304

Jermaine Jenas’ wife breaks silence over ‘difficult time’ following his texting scandal

    0
    Jermaine Jenas’ wife breaks silence over ‘difficult time’ following his texting scandal


    Ellie Jenas – wife of former footballer and BBC presenter Jermaine Jenas – has spoken publicly for the first time about the challenging period she and her family have endured following Jermaine’s texting scandal.

    The 41-year-old was dismissed from the BBC earlier this year after he was accused of sending ‘inappropriate’ texts to two female colleagues.

    Appearing at the launch of her new children’s skincare brand, Preppy, Ellie admitted that the past three months had been “incredibly difficult.”

    Notably, she was seen without her wedding ring at the event.

    Ellie Jenas admitted that things have been difficult following Jermaine’s dismissal. (Credit: SplashNews)

    Ellie Jenas’ focus has been on their children

    Speaking at the Battersea launch, Ellie opened up about the strain the scandal had placed on her and her family.

    “It has been an incredibly hard and difficult time for myself and my family.” She began. “My only focus has been on our children and building my brand Preppy.

    Now I just want to move forward. Naturally, this has been hard for me, but my only concern has been our children, and that will not change.

    Despite the challenges, Ellie gushed about the brand.

    She explained that she has been working on it for three years and is “incredibly proud” of the products.

    Ellie has been married to Jermaine since 2011 and shares three children with him.

    During the launch, the former model declined to discuss the state of their marriage.

    However, she acknowledged that Jermaine has been supportive of her business venture.

    “Jermaine is excited for the business and what it holds for me in the future. He has been supportive of me in this respect and as a working mum.” She shared.

    The couple’s three children have played a big role in the development of Preppy.

    Ellie revealed that they were the inspiration behind the brand, which focuses on self-care products for older children and teens.

    “Our daughter thought up the name Preppy.” She revealed.

    Jermaine Jenas was fired from his broadcasting job after being accused of 'inappropriate' behaviour.

    Jermaine Jenas was fired from his broadcasting job after being accused of ‘inappropriate’ behaviour. (Credit: SplashNews)

    Jermaine Jenas apologised to his wife over the texting scandal

    Ellie and her business partner, Steph Leggit, launched Preppy with a range of 15 products, including body wash, deodorant, and shower whip.

    The brand was developed from Ellie’s experience as a mother of three.

    “We realised no one was making products for older children and more often than not they steal mine,” Ellie explained.

    The products are already available online and will debut on TikTok Shop, with plans for a nationwide retail rollout in the new year.

    Jermaine Jenas previously addressed the scandal in an emotional statement, where he apologised to Ellie and acknowledged his mistake.

    Speaking to The Sun in August, he said: “I’m the married one – I’m at fault. The overriding panic by a mile was that I could lose my family.

    “Ellie is absolutely raging. She’s human. When I see her now, she is just making sure the children are okay.”

    While Jermaine insisted his actions were “unwise” rather than illegal, he apologised for making anybody feel “uncomfortable”.

    Although the future of Ellie and Jermaine’s marriage is unclear, Ellie is determined to focus on her children and her growing business.

    “Preppy has been our baby for three years.” She gushed. “This has been keeping me busy too as all the children are at school now. Before I could go to everything but now I am juggling the children and meetings.

    It’s just brilliant and I’m enjoying it.”

    Read more: Jermaine Jenas details ‘difficult period’ in new video message amid texting scandal

    'I'm not happy': Jermaine Jenas sacked by BBC while live on air at talkSPORT

    So what do you think? Tell us on our Facebook page @EntertainmentDailyFix.



    Source link

    Engineering puzzle game Poly Bridge 3 gets Steam Deck Verified with a major update

    0
    Engineering puzzle game Poly Bridge 3 gets Steam Deck Verified with a major update


    Highly rated, and now you can take it happily on the go, Poly Bridge 3 had a big upgrade and it’s now Steam Deck Verified. It already had Native Linux support since release.

    The 1.5.0 update brings with it Faulty Drive, a whole new world to play through. The developer said it’s a “three-sheep difficulty world that takes place in cyberspace” and so to “Expect the unexpected in a world where anything can happen”.

    What else is new:

    Gamepad Support




    Enjoy Poly Bridge 3 with full gamepad support on a Steam Deck or on the couch in Steam Big Picture Mode!
    New snap-to-node functionality (gamepad only) to make drawing easier!
    Check out the new Gamepad options to customize controls to your liking.

    Curated Gallery




    Added a new “Curated Gallery” option to ensure replays are appropriate for all ages.
    Curated replays are hand-picked by the dev team and cover a wide range of solutions.

    Polish and Bug Fixing




    Lots of small polish and bug fixing across the game

    Original launch trailer below:

    Writing on Twitter, the developer mentioned: “Got Poly Bridge 3 verified for Steam Deck. Fair bit of work since game was designed for keyboard/mouse. Will build controller support in from the start for all future projects.”

    Poly Bridge 3 | Release Date: 30th May 2023

    Official links and where to buy from:

    Article taken from GamingOnLinux.com.



    Source link

    Assassin’s Creed Syndicate gets surprise 60 FPS update on consoles | TheSixthAxis

    0
    Assassin’s Creed Syndicate gets surprise 60 FPS update on consoles | TheSixthAxis


    Almost a decade after the game’s original release, Assassin’s Creed Syndicate has been patched on modern consoles with a significant performance update.

    Ubisoft has confirmed that the underrated, Victorian era stealth ’em up now runs at a silky smooth 60 frame per second on PlayStation 5 and Xbox Series X|S. There’s also the option to simultaneously crank the game’s resolution to 4K though this won’t be available to those playing on the slightly less robust Xbox Series S.

    The publisher didn’t give a specific reason for Syndicate’s mini makeover, nor has it hinted at similar updates for other popular AC titles including the Ezio Collection, Unity, and Black Flag. With Assassins’ Creed Shadows pushed back to February 2025, it likely wanted to drum up interest in older entries with many players having overlooked Syndicate.

    Assassin’s Creed Syndicate launched in October 2015 and although it was met with generally positive reviews, fans had become overfamiliar with the franchise’s time-hopping, open world formula even if introduced dual protagonists and a genuinely interesting setting. This would be the final instalment before Ubisoft hit the reset button, transporting players back to Ancient Egypt in 2017’s Assassin’s Creed: Origins. Its debut drew a line in the sand and would serve as the foundation for future titles, treading further and further into action RPG territory.

    Last year’s Assassin’s Creed Mirage went against the grain with a smaller, more focused adventure that dialled back many of those roleplaying systems. In our review we scored it a fair 7/10, Aran saying: “Assassin’s Creed Mirage will appeal to anyone who’s been pining for a return to the old school open world stealth of the earlier games. It’s pretty much exactly that with a few extra refinements and additions. Some of those additions are a bit distracting and immersion breaking, but nothing gets in the way of some good old fashioned assassinations.”

    Source: Ubisoft



    Source link

    Bitcoin tests $94K

    0
    Bitcoin tests K


    The cryptocurrency market continues its rapid ascent, with Bitcoin, the leading cryptocurrency, reaching an all-time high of $94,000 yesterday evening. The surge in the market began with the US elections and shows no signs of slowing down.

    Bitcoin Surges to New Heights

    According to CoinMarketCap data, Bitcoin peaked at $94,020 before stabilizing around $92,500 following profit-taking at the higher levels. This bullish momentum has significantly impacted the total cryptocurrency market cap, which now exceeds $3 trillion, marking a new milestone.

    Trump’s Media Company Eyeing the Crypto Market

    The recent rally was fueled by speculation that Donald Trump’s media company may soon enter the cryptocurrency space. Reports from the Financial Times suggest that Trump Media and Technology Group, the operator of Truth Social, is close to acquiring Bakkt, a cryptocurrency company backed by the NYSE-owned Intercontinental Exchange.

    With Trump expected to take a crypto-friendly stance in his anticipated second term, the market is optimistic about further growth and adoption.

    Bitcoin ETF Trading Boosts Market Confidence

    Adding to the market’s momentum, Bitcoin ETF options began trading on Nasdaq yesterday, generating approximately $2 billion in trading volume on the first day. The approval of Spot Bitcoin ETFs earlier this year has significantly enhanced institutional interest, driving market confidence.

    Fear and Greed Index Shows “Extreme Greed”

    The Fear and Greed Index, a tool that measures investor sentiment in the cryptocurrency market, surged to 90, indicating “extreme greed.” This reflects the growing optimism and robust demand in the market, as investors continue to pour funds into cryptocurrencies.

    Key Highlights:

    Bitcoin’s Record Surge: Reached $94,000, stabilizing at $92,500.

    Market Cap Milestone: The total crypto market cap surpassed $3 trillion.

    Trump’s Crypto Moves: Speculation around Trump Media acquiring Bakkt.

    ETF Trading Impact: Bitcoin ETF options trading on Nasdaq boosts confidence.

    Investor Sentiment: The Fear and Greed Index hits 90, signaling bullish trends.

    The cryptocurrency market remains on an upward trajectory, with Bitcoin leading the way as institutional interest and market adoption continue to grow.

    You May Also Like

    Follow us on TWITTER (X) and be instantly informed about the latest developments…

    Copy URL



    Source link

    I’m A Celebrity: Barry McGuigan misses out on luxury item

      0
      I’m A Celebrity: Barry McGuigan misses out on luxury item


      I’m A Celebrity viewers were emotional last night as Barry McGuigan missed out on getting his luxury item.

      It came after Monday’s episode when former boxer Barry broke down in tears over his daughter’s death. Barry’s daughter, Danika, died in 2019 from bowel cancer at the age of just 33.

      On Tuesday night’s show (November 20), the campmates had the chance to win their luxury items from home. But Barry and Jane Moore were the only celebs not to receive their items.

      Alan and Danny tried to win the campmates their luxury items (Credit: ITV)

      Barry McGuigan on I’m A Celebrity

      During the ep, Danny Jones and Alan Halsall took on a jungle challenge in a bid to win the luxury packages.

      They were told that one at a time, a parcel would appear on the conveyor belt in front of them. Each parcel represented a luxury item for the campmates.

      However, they had to answer a question about their campmates correctly to ensure the box didn’t fall off the belt at the end.

      They managed to win eight out of the 10 boxes. Barry and Loose Women star Jane’s luxury items weren’t won, leaving Danny and Alan feeling gutted.

      Barry McGuigan talking in the I'm A Celebrity camp

      Barry didn’t get his item (Credit: ITV)

      Later in the show, the campmates sat around the campfire as they opened their items. Among the items were GK Barry‘s dressing gown, Coleen Rooney‘s personalised pillow with a family photo on it and Tulisa‘s inflatable chair.

      However, Barry and Jane watched on without receiving their items.

      Aww petition for Barry to have his luxury item please.

      Barry praised Danny and Alan and thanked them for trying to win all the items. However, on X, viewers were heartbroken for him and begged bosses to give him the item following his tears this week.

      One person said: “Aww petition for Barry to have his luxury item please,” followed by tearful eye emojis.

      Jane Moore with her hand on her face on I'm A Celebrity

      Jane also didn’t get her item (Credit: ITV)

      Another wrote: “My heart aches for Barry, he’s so wholesome, calling a petition for him to receive his package!!”

      Someone else added: “Barry makes my heart ache. He’s such a lovely man. Please give him his gift or I’ll cry.”

      Another said: “Heartbroken for Barry not getting his luxury item after opening up about his daughter’s death last night.”

      Others were baffled over the stars getting their items just days in. One said: “Why are they getting luxury items? Seems a bit ridiculous!!”

      YouTube video player

      Read more: I’m A Celebrity host Ant McPartlin left ‘crying’ as Barry McGuigan breaks down over daughter’s death

      What did you think of the show last night? Let us know by leaving a comment on our Facebook page @EntertainmentDailyFix.



      Source link

      Facebook’s Vision of the Metaverse: A New Digital Frontier – Web3oclock

      0
      Facebook’s Vision of the Metaverse: A New Digital Frontier – Web3oclock


      What is the Metaverse According to Meta?

      Key Pillars of Meta’s Metaverse Initiatives

      The Technologies Behind Meta’s Metaverse

      Potential Challenges and Criticisms

      1. Horizon Worlds:

      2. Horizon Workrooms:

      3. Horizon Venues:

      4. Oculus and Next-Gen VR/AR Hardware:

      5. Meta Avatars and AI-Powered Personalization:

      6. Spark AR and Augmented Reality Initiatives:

      7. Virtual Economy and Digital Commerce:

      8. Project Cambria (High-End VR Headset):

      The Technologies Behind Meta’s Metaverse:

      1. Virtual Events: 

      Potential Challenges and Criticisms of Meta:



      Source link

      NBA Star Shaquille O’Neal to Pay $11 Million in Astrals NFT Lawsuit Settlement – Cryptoflies News

      0
      NBA Star Shaquille O’Neal to Pay  Million in Astrals NFT Lawsuit Settlement – Cryptoflies News


      0

      Former basketball player Shaquille O’Neal has agreed to settle a lawsuit related to the Astrals non-fungible token (NFT) project, committing to pay $11 million. The settlement awaits court approval, which would conclude the legal dispute tied to O’Neal’s involvement in the project.

      The lawsuit began last year when investors in the Astrals project accused O’Neal of promoting and selling unregistered securities. 

      The case revolved around Astrals NFTs, described as 10,000 “metaverse-ready 3D avatars,” and the associated Galaxy token. 

      While O’Neal was not directly in charge of the project, a Florida court ruled in August that he played the role of a “seller,” linking him to its promotion.

      Legal issues surrounding NFTs are not new, with other high-profile figures and companies facing similar accusations.

      You Might Be Interested In

      In June, Dapper Labs settled a lawsuit over its NBA Top Shot NFTs for $4 million. In July, a U.S. judge rejected DraftKings’ effort to dismiss a class action claiming its NFTs were unregistered securities. Shortly after, DraftKings shut down its Reignmakers NFT game and marketplace, citing legal risks.

      Last November, Cristiano Ronaldo was sued over his NFT collaboration with Binance. 

      Last year, media company Impact Theory paid a $6.1 million penalty for issuing unregistered NFT securities, while the creators of Stoner Cats NFTs were fined $1 million for similar violations.

      Regulatory action has continued this year. Earlier this month, the SEC issued a Wells notice to Immutable, suggesting legal action might follow over alleged securities law violations tied to its IMX token. The SEC also targeted OpenSea, raising concerns about unregistered securities sold on its platform.



      Source link

      MetaMask Releases Extension V12.6, Introducing Gas Station And Chain Permission Controls

      0
      MetaMask Releases Extension V12.6, Introducing Gas Station And Chain Permission Controls


      In Brief

      MetaMask has released its v12.6 extension, introducing Gas Station for token swaps without requiring ETH for gas fees, and Chain Permission controls that provide users with control over account and network permissions.

      MetaMask Releases Extension V12.6, Introducing Gas Station And Chain Permission Controls

      Self-custodial cryptocurrency wallet MetaMask announced the release of extension v12.6, introducing a new feature that allows users to swap tokens without needing ETH to cover gas fees.

      The update introduces the Gas Station feature, which becomes accessible when users enable Smart Transactions with MetaMask Swaps. This functionality simplifies and accelerates transactions by incorporating network fees directly into the quotes users receive. Furthermore, users are not restricted to a single liquidity source for their Swaps. Instead, the platform continues to aggregate data from a wide range of providers, including decentralized exchange aggregators, market makers, and decentralized exchanges (DEXs), ensuring competitive pricing and minimal network fees.

      Currently, the Gas Station feature is available on the MetaMask Extension via the Ethereum mainnet, with plans to expand to mobile devices in the near future. In order to use it, the swap value must cover the cost of gas and include supported assets such as USDT, USDC, DAI, ETH, wETH, wBTC, wstETH, and wSOL.

      Chain Permission Controls Enable More Granular Account And Network Management

      Another update includes the ability for users to sign transactions on any permitted network without the need to manually approve network switches. This enhancement is made possible through the Chain Permission controls, which provide users with more detailed control over account and network permissions.

      Permissions are granted to each decentralized application (dApp) at the time of connection, offering users flexibility and oversight. When a user connects to a dApp, the account in use and all default-enabled networks automatically gain permission for that dApp. However, users can modify these permissions either during the initial connection process or afterward through the All Permissions section. For those managing their own infrastructure, such as home stakers, the update also allows users to review and adjust their chosen RPC URLs as needed.

      This feature not only enhances MetaMask‘s usability but also strengthens its security framework. By enabling users to limit dApp permissions, it reduces potential exposure to security risks while maintaining a seamless user experience.

      MetaMask is a free digital wallet tailored for managing ETH and Ethereum-based tokens, acting as a gateway to the Ethereum blockchain. It allows users to perform a range of functions, including buying, sending, receiving, and swapping cryptocurrencies and NFTs. Additionally, it offers support for adding custom networks, extending its usability beyond the Ethereum ecosystem.

      Disclaimer

      In line with the Trust Project guidelines, please note that the information provided on this page is not intended to be and should not be interpreted as legal, tax, investment, financial, or any other form of advice. It is important to only invest what you can afford to lose and to seek independent financial advice if you have any doubts. For further information, we suggest referring to the terms and conditions as well as the help and support pages provided by the issuer or advertiser. MetaversePost is committed to accurate, unbiased reporting, but market conditions are subject to change without notice.

      About The Author


      Alisa, a dedicated journalist at the MPost, specializes in cryptocurrency, zero-knowledge proofs, investments, and the expansive realm of Web3. With a keen eye for emerging trends and technologies, she delivers comprehensive coverage to inform and engage readers in the ever-evolving landscape of digital finance.

      More articles


      Alisa Davidson










      Alisa, a dedicated journalist at the MPost, specializes in cryptocurrency, zero-knowledge proofs, investments, and the expansive realm of Web3. With a keen eye for emerging trends and technologies, she delivers comprehensive coverage to inform and engage readers in the ever-evolving landscape of digital finance.








      More articles



      Source link

      Grayscale to Launch Bitcoin ETF Options Following BlackRock’s Record Debut – Decrypt

      0
      Grayscale to Launch Bitcoin ETF Options Following BlackRock’s Record Debut – Decrypt



      Crypto asset manager Grayscale Investments plans to roll out options trading on its spot Bitcoin ETFs on Wednesday amid the first glimpses of solid investor appetite for such products.

      The announcement comes a day after BlackRock’s iShares Bitcoin Trust (IBIT) achieved record-breaking activity on its first day of options trading, pushing Bitcoin to a new all-time high.

      Grayscale will launch options trading on GBTC (Grayscale Bitcoin Trust) and BTC (Bitcoin Mini Trust) to “further [develop] the ecosystem around our US-listed Bitcoin ETPs,” it said.

      Following the Options Clearing Corporation’s (OCC) approval of Bitcoin ETF options, Grayscale quickly filed an updated prospectus for its Bitcoin Covered Call ETF on January 11.

      The tooling aims to generate income by employing a covered call strategy—writing and buying options contracts on Bitcoin exchange-traded products (ETPs) while holding Bitcoin or GBTC as collateral.

      Bloomberg ETF analyst Seyffart called attention to the speed of Grayscale’s response following the OCC’s clearance, tweeting Tuesday that the asset manager was “wasting no time.”

      “They’ve filed an updated prospectus for their Bitcoin Covered Call ETF,” Seyffart tweeted. “The fund will offer exposure to $GBTC & $BTC while writing &/or buying options contracts on Bitcoin ETPs for income.”

      Grayscale follows the unprecedented debut of BlackRock’s IBIT options, which recorded nearly $1.9 billion in notional exposure traded on its first day

      Seyffart shared details on X, observing that 354,000 contracts were exchanged, including 289,000 calls and 65,000 puts, representing a 4.4:1 call-to-put ratio.

      The ratio indicates that a significantly larger number of investors placed bets on Bitcoin’s price rise (calls) compared to those hedging against a potential price drop (puts). 

      “These options were almost certainly part of the move to the new Bitcoin all-time highs today,” Seyffart wrote, referring to Bitcoin’s surge to $94,041 on Tuesday.

      Bloomberg’s senior ETF analyst Eric Balchunas characterized the $1.9 billion trading volume as “unheard of” for any given options trading within an ETF during its first day.

      “$For context, BITO did $363 million, and that’s been around for four years,” Balchunas wrote on X, referring to ProShares’ futures Bitcoin ETF.

      Roughly 73,000 options contracts were traded in the first 60 minutes, placing IBIT among the top 20 most active non-index options on its opening day.

      Grayscale’s launch comes a year after its major legal victory against the SEC. Last August, the U.S. Court of Appeals ordered the SEC to revisit its denial of Grayscale’s application to convert its Bitcoin Trust into a spot ETF.

      This ruling was a turning point for crypto ETFs, challenging regulatory resistance that had stalled their approvals for nearly a decade.

      Edited by Sebastian Sinclair

      Daily Debrief Newsletter

      Start every day with the top news stories right now, plus original features, a podcast, videos and more.



      Source link

      Step-by-Step Guide to Creating an LLM-Based App for Chat with Papers

      0
      Step-by-Step Guide to Creating an LLM-Based App for Chat with Papers


      Staying updated with the latest in machine learning (ML) research can feel overwhelming. With the steady stream of papers on large language models (LLMs), vector databases, and retrieval-augmented generati on (RAG) systems, it’s easy to fall behind. But what if you could access and query this vast research library using natural language? In this guide, we’ll create an AI-powered assistant that mines and retrieves information from Papers With Code (PWC), providing answers based on the latest ML papers.

      Our app will use a RAG framework for backend processing, incorporating a vector database, VertexAI’s embedding model, and an OpenAI LLM. The frontend will be built on Streamlit, making it simple to deploy and interact with.

      Step 1: Data Collection from Papers With Code

      Papers With Code is a valuable resource that aggregates the latest ML papers, source code, and datasets. To automate data retrieval from this site, we’ll use the PWC API. This allows us to collect papers related to specific keywords or topics.

      Retrieving Papers Using the API

      To search for papers programmatically:

      Access the PWC API Swagger UI and locate the papers/ endpoint.

      Use the q parameter to enter keywords for the topic of interest.

      Execute the query to retrieve data.

      Each response includes the first set of results, with additional pages accessible via the next key. To retrieve multiple pages, you can set up a function that loops through all pages based on the initial result count. Here’s a Python script to automate this:

      import requests
      import urllib.parse
      from tqdm import tqdm

      def extract_papers(query: str):
      query = urllib.parse.quote(query)
      url = f”https://paperswithcode.com/api/v1/papers/?q={query}
      response = requests.get(url).json()
      count = response[“count”]
      results = response[“results”]

      num_pages = count // 50
      for page in tqdm(range(2, num_pages)):
      url = f”https://paperswithcode.com/api/v1/papers/?page={page}&q={query}
      response = requests.get(url).json()
      results.extend(response[“results”])
      return results

      query = “Large Language Models”
      results = extract_papers(query)
      print(len(results))

      Formatting Results for LangChain Compatibility

      Once extracted, convert the data to LangChain-compatible Document objects. Each document will contain:

      page_content: stores the paper’s abstract.

      metadata: includes attributes like id, arxiv_id, url_pdf, title, authors, and published.

      from langchain.docstore.document import Document

      documents = [
      Document(
      page_content=result[“abstract”],
      metadata={
      “id”: result.get(“id”, “”),
      “arxiv_id”: result.get(“arxiv_id”, “”),
      “url_pdf”: result.get(“url_pdf”, “”),
      “title”: result.get(“title”, “”),
      “authors”: result.get(“authors”, “”),
      “published”: result.get(“published”, “”)
      },
      )
      for result in results
      ]

      Chunking for Efficient Retrieval

      Since LLMs have token limitations, breaking down each document into chunks can improve retrieval and precision. Using LangChain’s RecursiveCharacterTextSplitter, set chunk_size to 1200 characters and chunk_overlap to 200. This will generate manageable text chunks for optimal LLM input.

      text_splitter = RecursiveCharacterTextSplitter(
      chunk_size=1200,
      chunk_overlap=200,
      separators=[“.”]
      )
      splits = text_splitter.split_documents(documents)
      print(len(splits))

      Step 2: Creating an Index with Upstash

      To store embeddings and document metadata, set up an index in Upstash, a serverless database ideal for our project. After logging into Upstash, set your index parameters:

      Region: closest to your location.

      Dimensions: 768, matching VertexAI’s embedding dimension.

      Distance Metric: cosine similarity.

      Then, install the upstash-vector package:

      pip install upstash-vector

      Use the credentials generated by Upstash (URL and token) to connect to the index in your app.

      from upstash_vector import Index

      index = Index(
      url=“<UPSTASH_URL>”,
      token=“<UPSTASH_TOKEN>”
      )

      Step 3: Embedding and Indexing Documents

      To add documents to Upstash, we’ll create a class UpstashVectorStore which embeds document chunks and indexes them. This class will include methods to:

      from typing import List, Optional, Tuple, Union
      from uuid import uuid4
      from langchain.docstore.document import Document
      from langchain.embeddings.base import Embeddings
      from tqdm import tqdm
      from upstash_vector import Index

      class UpstashVectorStore:
      def __init__(self, index: Index, embeddings: Embeddings):
      self.index = index
      self.embeddings = embeddings

      def add_documents(
      self,
      documents: List[Document],
      batch_size: int = 32
      ):

      texts, metadatas, all_ids = [], [], []

      for document in tqdm(documents):
      texts.append(document.page_content)
      metadatas.append({“context”: document.page_content, **document.metadata})

      if len(texts) >= batch_size:
      ids = [str(uuid4()) for _ in texts]
      all_ids += ids
      embeddings = self.embeddings.embed_documents(texts)
      self.index.upsert(vectors=zip(ids, embeddings, metadatas))
      texts, metadatas = [], []

      if texts:
      ids = [str(uuid4()) for _ in texts]
      all_ids += ids
      embeddings = self.embeddings.embed_documents(texts)
      self.index.upsert(vectors=zip(ids, embeddings, metadatas))
      print(f”Indexed {len(all_ids)} vectors.”)
      return all_ids

      def similarity_search_with_score(
      self, query: str, k: int = 4
      ) -> List[Tuple[Document, float]]:

      query_embedding = self.embeddings.embed_query(query)
      results = self.index.query(query_embedding, top_k=k, include_metadata=True)
      return [(Document(page_content=metadata.pop(“context”), metadata=metadata), score)
      for metadata, score in results]

      To execute this indexing:

      from langchain.embeddings import VertexAIEmbeddings

      embeddings = VertexAIEmbeddings(model_name=“textembedding-gecko@003”)
      upstash_vector_store = UpstashVectorStore(index, embeddings)
      ids = upstash_vector_store.add_documents(splits, batch_size=25)

      Step 4: Querying Indexed Papers

      With the abstracts indexed in Upstash, querying becomes straightforward. We’ll define functions to:

      Retrieve relevant documents.

      Build a prompt using these documents for LLM responses.

      def get_context(query, vector_store):
      results = vector_store.similarity_search_with_score(query)
      return “\n===\n”.join([doc.page_content for doc, _ in results])

      def get_prompt(question, context):
      template = “””
      Use the provided context to answer the question accurately.

      %CONTEXT%
      {context}

      %Question%
      {question}

      Answer:
      “””
      return template.format(question=question, context=context)

      For example, if you ask about the limitations of RAG frameworks:

      query = “What are the limitations of the Retrieval Augmented Generation framework?”
      context = get_context(query, upstash_vector_store)
      prompt = get_prompt(query, context)

      Step 5: Building the Application with Streamlit

      To make our app user-friendly, we’ll use Streamlit for a simple, interactive UI. Streamlit makes it easy to deploy ML-powered web apps with minimal code.

      import streamlit as st
      from langchain.chat_models import AzureChatOpenAI

      st.title(“Chat with ML Research Papers”)
      query = st.text_input(“Ask a question about ML research:”)

      if st.button(“Submit”):
      if query:
      context = get_context(query, upstash_vector_store)
      prompt = get_prompt(query, context)
      llm = AzureChatOpenAI(model_name=“<MODEL_NAME>”)
      answer = llm.predict(prompt)
      st.write(answer)

      Benefits and Limitations of Retrieval-Augmented Generation (RAG)

      RAG systems offer unique advantages, especially for ML researchers:

      Access to Up-to-Date Information: RAG lets you pull information from the latest sources.

      Enhanced Trust: Answers grounded in source documents make results more reliable.

      Easy Setup: RAGs are relatively straightforward to implement without needing extensive computing resources.

      However, RAG isn’t perfect:

      Data Dependence: RAG accuracy hinges on the data fed into it.

      Not Always Optimal for Complex Queries: While fine for demos, real-world applications may need extensive tuning.

      Limited Context: RAG systems are still limited by the LLM’s context size.

      Conclusion

      Building a conversational assistant for machine learning research using LLMs and RAG frameworks is achievable with the right tools. By using Papers With Code data, Upstash for vector storage, and Streamlit

      for a user interface, you can create a robust application for querying recent research.

      Further Exploration Ideas:

      Use the full paper text rather than just abstracts.

      Experiment with metadata filtering to improve precision.

      Explore hybrid retrieval techniques and re-ranking for more relevant results.

      Whether you’re an ML enthusiast or a researcher, this approach to interacting with research papers can save time and streamline the learning process.



      Source link

      Popular Posts

      My Favorites