Add SetFit model

README.md

@@ -6,21 +6,225 @@ tags:
 - text-classification
 - generated_from_setfit_trainer
 base_model: sentence-transformers/paraphrase-mpnet-base-v2
-datasets:
-- ethos
 metrics:
 - accuracy
 widget:
-- text: Why are niggers superior athletically to all other races?
-- text: An ugly man in a wig and dress, is still just an ugly man in a wig and dress...
-    NOT A WOMAN!!!
-- text: Perfect place for Lana- China and Myanmar. No trouble there. They know how
-    to handle those pigs.
-- text: No counting when Hispanics get probation if they do find them self in some
-    trouble they get probation then get deported and if they come back boom there
-    in violation of the probation we gave them but knowingly sent them back. Those
-    rats are smart. Bunch of GOVERNMENT BULL SHIT
-- text: Most Americans are nazis or kkk members so they elected  Trump.
+- text: ' CercleFinance.com KKR and The Global Atlantic Financial have announced the
+    finalization of their transaction. KKR completed the acquisition of the remaining
+    37 of Global Atlantic bringing its stake to 100 . ''Since day one Global Atlantic
+    has been a great fit for KKR both from a business and cultural perspective. We
+    look forward to collaborating even more closely with Global Atlantic so that we
+    can realize more of the synergies we have discovered over the first three years
+    of our strategic partnership '' said Joseph Bae and Scott Nuttall Co. Chief Executive
+    Officers of KKR. ''KKR and Global Atlantic are a powerful combination. Our shared
+    culture and commitment to excellence continue to enhance our ability to think
+    and invest long term and deliver compelling solutions to our customers and policyholders
+    '' said Allan Levine Co Founder Chairman and CEO from Global Atlantic. Global
+    Atlantic will continue to be led by its management team and operate under the
+    Global Atlantic brand. Copyright 2024 CercleFinance.com. All rights reserved.
+    Did you like this article ? Share it with your friends using the buttons below.
+    Source link 85 '
+- text: 'As we move toward a society where anyone can buy Bitcoin major financial
+    institutions around the world are working to develop next generation business
+    infrastructure with an eye toward a society in which digital financial assets
+    tokenized on the blockchain are exchanged. proceeding. Nomura Holdings is one
+    such company. CoinDesk a major Web3 media company had started its own project
+    before launching its Japanese version in 2019. Over the past five years when the
+    crypto asset virtual currency market has experienced a boom and bust period Nomura
+    has been building a next generation business foundation within the company that
+    will lead to the current digital company. Hajime Ikeda is the key person leading
+    this digital company. In 2023 the market was so cold that it was called Crypto
+    Winter but the world''s attention was focused on RWA tokenization initiatives
+    being promoted by major financial institutions in Europe the United States Japan
+    and Singapore. RWA is an abbreviation that stands for Real World Asset and unlike
+    crypto assets such as Bitcoin it refers to financial assets that actually exist.
+    In other words if assets such as real estate bonds fiat currency gold and renewable
+    energy are tokenized on the chain the market size will expand to trillions or
+    even tens of trillions of dollars. Boston Consulting estimates that the RWA market
+    could grow to as much as 16 trillion by 2030. How will the infrastructure human
+    resources and tools that Nomura''s digital company has built up function in this
+    new expanding market? We spoke to Mr. Ikeda about the near future while looking
+    at the current status of major projects undertaken by Digital Company. Komainu
+    Mr. Ashley behind the birth of the project The service that manages financial
+    assets held by institutional investors is called custody and this is the custody
+    business that Nomura first embarked on when building a business foundation for
+    digital assets. In 2018 Nomura launched the project Komainu in collaboration with
+    European companies Ledger and CoinShares. Custody is an essential service for
+    the functioning of current financial markets. For example the Bank of New York
+    now Bank of New York Mellon founded in the 1700s by Alexander Hamilton the author
+    of the U.S. Constitution is the oldest bank in the United States and has been
+    providing custody services for the securities of institutional investors for over
+    200 years. If a bank were to suddenly stop its operations the U.S. financial market
+    the world''s largest would be paralyzed. How should Nomura respond to digital
+    assets? After many heated discussions within the company Nomura decided to co
+    found Komainu. The person who had a major influence on his decision was Stephen
+    Ashley who was head of wholesale at the time. Hajime Ikeda Executive Officer Digital
+    Company and Marketing Sales Division Nomura Holdings At that time there were various
+    opinions within the company regarding digital assets. Many events and problems
+    including hacking occurred in society relating to digital assets . However the
+    market for digital assets did not grow. Custody will definitely be necessary Mr.
+    Ikeda recalled at the time. We started from here and have continued to build on
+    digital assets. In North America JPMorgan Chase the largest US bank was recently
+    conducting a pilot project to use Quorum an independently developed blockchain
+    based on Ethereum to perform payments and remittances in tokenized fiat currency.
+    JP Morgan sold Quorum to ConsenSys the developer of the wallet Metamask in 2020.
+    At the same time it acquired a portion of ConsenSys stock. Currently JP Morgan
+    through its subsidiary Onyx is developing a project for corporations that utilizes
+    tokens. In Japan in 2018 an incident occurred in which crypto assets worth 58
+    billion yen were hacked and stolen from the crypto asset exchange company Coincheck.
+    A report by the US research firm Chainalysis later revealed that the North Korean
+    hacking organization Lazarus was involved in the incident. Laser Digital The world''s
+    giant asset management company on the rise In 2021 when it was reported that major
+    US investment bank Goldman Sachs was planning to open a crypto asset trading desk
+    Nomura was preparing to launch Laser Digital. The following year Laser established
+    its headquarters in Zurich Switzerland. The three pillars of its business are
+    digital asset management services and trading for institutional investors and
+    startup investment. Mr. Ashley who has been in charge of the wholesale division
+    has been appointed as Laser''s chairman. This shows how serious Nomura is about
+    this business. To date as part of its asset management services it has been operating
+    funds based on Bitcoin and Ethereum but for custody they have immediately started
+    using Comainu which they have been developing. Laser will launch a Japanese subsidiary
+    in October 2023 with Hideaki Kudo who has been leading digital asset business
+    development together with Ikeda as its head. Mr. Kudo joined Nomura Asset Management
+    after researching astrophysics and has led the development of financial products.
+    Laser Digital Japan Representative Director and President Hideaki Kudo After the
+    start of Komaine there was a certain level of understanding within the company
+    about this initiative laser says Ikeda. When starting a trading business 24 hour
+    monitoring is necessary. By supporting Asian time through Laser Digital Japan
+    and starting a 24 hour system there is a world that will become visible. Should
+    we do something? What can we do? It will become clear. U.S. asset management companies
+    that manage huge amounts of funds are also becoming more active in developing
+    financial services that handle crypto assets and tokenized RWA. One such company
+    is BlackRock which manages over 9 trillion in assets and has been an early strategic
+    investor in the tokenized RWA space. A typical example is the investment in the
+    American company Circle which issues the stablecoin USD Coin USDC which is linked
+    to the US dollar. In Circle''s 400 million funding round in 2022 Fidelity a major
+    asset management company like BlackRock participated. At that time BlackRock and
+    Circle entered into a strategic partnership. Additionally BlackRock has announced
+    plans to list a bitcoin spot exchange traded fund ETF in 2023 and has filed an
+    application for listing with the U.S. Securities and Exchange Commission SEC .
+    If the world''s largest asset management company starts operating a Bitcoin ETF
+    a huge amount of money will flow into the crypto asset market. Laser''s workforce
+    is expected to increase to around 100 people worldwide by 2024. Mr. Ikeda emphasizes
+    that by utilizing technology such as blockchain we can realize the customer journey
+    of digital financial products in areas that cannot be covered by traditional financial
+    services. Blockchain AI alternative ...We are focusing on these three things in
+    the digital asset business going forward. What is alternative Alternative in English
+    means alternative . Assets that are different from traditional assets such as
+    stocks and bonds. BOOSTRY Supporting the digital asset market Japan started developing
+    RWA tokenization earlier than Europe and the United States. Japanese financial
+    institutions including Nomura are issuing new asset classes that are tokenized
+    on blockchain based platforms and are being offered primarily to retail investors
+    in primary markets with secondary markets such as Osaka Digital Exchange ODX We
+    have been rapidly developing an ecosystem that allows for trading and distribution
+    in the market. This digital asset is called Security Token ST and in Japan it
+    refers to assets linked to real estate and corporate bonds issued by companies
+    and can be securitized and converted into small units by tokenizing it on the
+    blockchain. A society is about to emerge in which many individual investors can
+    trade freely. For example a part of specific real estate such as a hot spring
+    inn popular with inbound tourists a hotel in Kyoto or a luxury condominium in
+    Tokyo can be purchased as a security token. Investors can obtain benefits preferential
+    tickets that come with security tokens while enjoying a return of a few percent.
+    Individuals purchasing financial products may not directly feel that blockchain
+    is being used behind the scenes of the financial product or that encrypted tokens
+    are being traded. In fact the process of buying real estate security tokens is
+    a much similar experience to buying any other asset class. The platform for issuing
+    and distributing security tokens was developed by BOOSTRY a subsidiary launched
+    by Nomura in 2019. Currently SBI Holdings holds 10 and Japan Exchange Group holds
+    5 of the company. BOOSTRY Representative Director and CEO Toshinori Sasaki right
+    speaking at Digital Securities Forum 2023 The reason behind the development of
+    ST for individuals in Japan is that most of the personal financial assets which
+    have ballooned to 2 100 trillion yen are cash and deposits and there has been
+    no shift from savings to investment in the past. Mr. Ikeda said The development
+    of security tokens will help bring idle money into the financial market in the
+    Japanese financial world and many companies revitalize the Japanese financial
+    market and improve the balance of this country which is biased towards cash and
+    deposits. There was a strong desire to eliminate the stagnation he said explaining
+    one reason why Japan took the lead in developing tokenized RWA targeting individual
+    investors. Will Japan''s sleepy money be attracted to security tokens after 2024
+    and some of it will flow into this new digital asset market? In a state of deflation
+    holding cash may be the right investment move to some extent. However there are
+    signs that the economy is moving out of this economic environment says Ikeda.
+    He added In a society where consumption and investment activities are becoming
+    digitally seamless I think a new investment environment is being created.I think
+    the major money flows will change. As the security token market continues to expand
+    opportunities for individual investors to invest in new digital assets will increase
+    in addition to traditional financial assets. Mr. Ikeda said that in the near future
+    when individuals are trying to build an optimal asset portfolio they will need
+    more detailed financial education and investment advice than ever before. Interview
+    Text Shigeru Sato Photo Keisuke Tada The post appeared first on Our Bitcoin News
+    . '
+- text: 'The apex court this week has given an interim stay on a significant Delhi
+    High Court ruling in early 2023 that a tax residency certificate TRC is sufficient
+    for a foreign investor''s claim to pay zero or lower tax under treaties between
+    India and overseas jurisdiction like Mauritius Singapore and The Netherlands.
+    Mumbai Foreign portfolio investors offshore private equity houses and others who
+    had acquired stocks through the foreign direct investment route are keeping their
+    fingers crossed after a Supreme Court ruling that could embolden the Indian income
+    tax I T authorities to question their tax benefits. The apex court this week has
+    given an interim stay on a significant Delhi High Court ruling in early 2023 that
+    a tax residency certificate TRC is sufficient for a foreign investor''s claim
+    to pay zero or lower tax under treaties between India and overseas jurisdiction
+    like Mauritius Singapore and The Netherlands. TRC is given by authorities of these
+    offshore financial centres to FPI and FDI entities set up in such countries for
+    betting on Indian securities. The treaties with Mauritius and Singapore though
+    amended later still allow foreign investors to avoid tax on capital gains generated
+    by selling stocks which were bought before April 2017. The Delhi HC ruling on
+    a writ petition by Blackstone Capital Partners a Singapore firm had said that
+    the I T department cannot go behind the TRC to challenge treaty benefits on the
+    grounds that the foreign entity incorporated in the treaty country is just a shell
+    company with no substance. Though tax officials continued to slap notices to foreign
+    investors after the HC''s decision asking a slew of questions on their origin
+    and nature of presence in the tax havens the HC ruling on the validity of TRC
+    came in handy in contesting the stand of the tax office. Now if the SC eventually
+    confirms the stay in its final ruling it would have an unsettling effect on many
+    overseas investors and influence the course of ongoing litigations. '
+- text: 'Sony Bank is taking a backseat position in management of Sony Payments Services
+    SPSV with Blackstone taking a majority stake. An American alternative investment
+    firm headquartered in New York City Blackstone has acquired the majority stake
+    from Sony Bank which is a wholly owned subsidiary of the wider Sony Group SPSV
+    is one of Japan''s largest payments service providers having scaled up a strong
+    presence across the country since its foundation in 1995 as part of the aforementioned
+    Sony Group. SPSV has solidified a healthy market position and earned the trust
+    of customers as a high quality payment service provider said SPSV President CEO
+    and Representative Director Hidehiko Nakamura We believe this partnership with
+    Blackstone will boost SPSV''s capabilities through investments in IT and talent
+    to help accelerate its growth journey particularly at an exciting time of growth
+    for the electronic payment industry in Japan. Although it became independent as
+    a standalone business in 2006 Sony Group''s Sony Bank still maintained a strong
+    degree of control over SPSV''s direction via its majority stake. Following its
+    sale to Blackstone SPSV is moving further away from its founding parent group.
+    However Sony and Blackstone are also long standing partners and further cooperation
+    is highly likely. Kenichiro Yoshida Chairman and CEO Sony Group remarked For the
+    past 30 years SPSV has led Japan''s cashless evolution making payments safe and
+    secure for customers. We believe Blackstone a long standing partner of Sony Group
+    can help continue the legacy that SPSV has formed and support its next phase of
+    growth. Meanwhile Sony Bank''s President and CEO Keiji Minami reflected that changing
+    trends in global payments and banking such as the shift towards cashless payments
+    and general diversification in payment types require greater adaptability. We
+    believe that Blackstone is the best partner bringing a global perspective and
+    its expertise and network in the payment business he said. For Blackstone the
+    agreement marks a continuation of the fund''s investments in fintech and payments
+    but also its debut in the Japanese fintech space which the group believes poses
+    strong potential. The firm notes that the Japanese electronic card payment market
+    is the third largest in the world with market penetration of 9.1 whilst the country
+    is also home to a JPY 22.7trn USD 18.9bn . As with other major developed financial
+    markets Japan has probed various forms of fintech innovation in recent years including
+    CBDCs and the Metaverse . The country has also witnessed the aforementioned trends
+    of economic digitalisation and widespread adoption of cashless payments. Atsuhiko
+    Sakamoto Head of Private Equity Blackstone Japan said We are thrilled to invest
+    in SPSV one of Japan''s leading payment services providers and a well established
+    financial technology company and expand our Japan Private Equity portfolio in
+    good neighbourhoods sectors with strong secular growth. Digitisation of the economy
+    is a key trend around the world including Japan and SPSV is exceptionally positioned
+    to benefit with its sophisticated technology and robust customer base. We''re
+    committed to bringing our operational and technology expertise and scale to support
+    SPSV''s growth. '
+- text: 'NiSource Inc. NYSE NI completes the issuance of a 19.9 indirect equity interest
+    in NIPSCO to Blackstone Infrastructure Partners affiliate for 2.16 billion with
+    an additional equity commitment of 250 million. The investment aims to strengthen
+    NIPSCO''s financial foundation support sustainable long term growth and fund ongoing
+    capital requirements for energy transition and reindustrialization of the Midwest. '
 pipeline_tag: text-classification
 inference: false
 model-index:
@@ -30,18 +234,18 @@ model-index:
       type: text-classification
       name: Text Classification
     dataset:
-      name: ethos
-      type: ethos
+      name: Unknown
+      type: unknown
       split: test
     metrics:
     - type: accuracy
-      value: 0.46524064171123
+      value: 0.2753623188405797
       name: Accuracy
 ---
 
 # SetFit with sentence-transformers/paraphrase-mpnet-base-v2
 
-This is a [SetFit](https://github.com/huggingface/setfit) model trained on the [ethos](https://huggingface.co/datasets/ethos) dataset that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) as the Sentence Transformer embedding model. A OneVsRestClassifier instance is used for classification.
+This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) as the Sentence Transformer embedding model. A OneVsRestClassifier instance is used for classification.
 
 The model has been trained using an efficient few-shot learning technique that involves:
 
@@ -56,7 +260,7 @@ The model has been trained using an efficient few-shot learning technique that i
 - **Classification head:** a OneVsRestClassifier instance
 - **Maximum Sequence Length:** 512 tokens
 <!-- - **Number of Classes:** Unknown -->
-- **Training Dataset:** [ethos](https://huggingface.co/datasets/ethos)
+<!-- - **Training Dataset:** [Unknown](https://huggingface.co/datasets/unknown) -->
 <!-- - **Language:** Unknown -->
 <!-- - **License:** Unknown -->
 
@@ -71,7 +275,7 @@ The model has been trained using an efficient few-shot learning technique that i
 ### Metrics
 | Label   | Accuracy |
 |:--------|:---------|
-| **all** | 0.4652   |
+| **all** | 0.2754   |
 
 ## Uses
 
@@ -91,7 +295,7 @@ from setfit import SetFitModel
 # Download from the 🤗 Hub
 model = SetFitModel.from_pretrained("Ankit15nov/setfit-ethos-multilabel-example")
 # Run inference
-preds = model("Why are niggers superior athletically to all other races?")
+preds = model("NiSource Inc. NYSE NI completes the issuance of a 19.9 indirect equity interest in NIPSCO to Blackstone Infrastructure Partners affiliate for 2.16 billion with an additional equity commitment of 250 million. The investment aims to strengthen NIPSCO's financial foundation support sustainable long term growth and fund ongoing capital requirements for energy transition and reindustrialization of the Midwest. ")
 ```
 
 <!--
@@ -121,12 +325,12 @@ preds = model("Why are niggers superior athletically to all other races?")
 ## Training Details
 
 ### Training Set Metrics
-| Training set | Min | Median  | Max |
-|:-------------|:----|:--------|:----|
-| Word count   | 3   | 27.0625 | 463 |
+| Training set | Min | Median | Max  |
+|:-------------|:----|:-------|:-----|
+| Word count   | 2   | 590.5  | 2491 |
 
 ### Training Hyperparameters
-- batch_size: (16, 16)
+- batch_size: (4, 4)
 - num_epochs: (1, 1)
 - max_steps: -1
 - sampling_strategy: oversampling
@@ -146,10 +350,18 @@ preds = model("Why are niggers superior athletically to all other races?")
 ### Training Results
 | Epoch  | Step | Training Loss | Validation Loss |
 |:------:|:----:|:-------------:|:---------------:|
-| 0.0063 | 1    | 0.2956        | -               |
-| 0.3125 | 50   | 0.1114        | -               |
-| 0.625  | 100  | 0.1177        | -               |
-| 0.9375 | 150  | 0.0695        | -               |
+| 0.0018 | 1    | 0.4292        | -               |
+| 0.0893 | 50   | 0.0057        | -               |
+| 0.1786 | 100  | 0.2115        | -               |
+| 0.2679 | 150  | 0.0003        | -               |
+| 0.3571 | 200  | 0.0022        | -               |
+| 0.4464 | 250  | 0.0003        | -               |
+| 0.5357 | 300  | 0.0083        | -               |
+| 0.625  | 350  | 0.0043        | -               |
+| 0.7143 | 400  | 0.0038        | -               |
+| 0.8036 | 450  | 0.0014        | -               |
+| 0.8929 | 500  | 0.0031        | -               |
+| 0.9821 | 550  | 0.0014        | -               |
 
 ### Framework Versions
 - Python: 3.10.14

config_setfit.json

@@ -1,4 +1,4 @@
 {
-  "labels": null,
-  "normalize_embeddings": false
+  "normalize_embeddings": false,
+  "labels": null
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74b1876698698a897b5dde2aed2384e43bc818e9b08ace8fab6f95a4b887ca87
+oid sha256:bb3404b7e1dcc5f929cbbf862d4027a7975055dce62f4cbfb379343b24072935
 size 437967672

model_head.pkl

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dbe8124353b411764e0ae10b625d49e511469a52288fb7137628b6c03f4e5056
-size 52836
+oid sha256:ccc9bc06e7df7f818b7ac25285e646a3e1d6c27a6168e14b3b47564ab88e0941
+size 46356