license: apache-2.0
language:
- en
library_name: transformers
metrics:
- accuracy
tags:
- crypto
- bitcoin
- news
- eth
- transformers
widget:
- text: >-
Bitcoin Vault (BTCV) traded 5.6% higher against the <mask> during the
twenty-four hour period ending at 14:00 PM Eastern on October 7th. In the
last week, Bitcoin Vault has traded down 2.7% against the dollar. One
Bitcoin Vault coin can now be bought for approximately $2.48 or 0.00012763
BTC on major cryptocurrency exchanges. Bitcoin Vault has a total market
cap of $5.20 million and approximately $63,451.00 worth of Bitcoin Vault
was traded on exchanges in the last day. Here's how other cryptocurrencies
have performed in the last day: Bitcoin (BTC)
example_title: MLM 1
- text: >-
Good morning. Here's what's <mask>:Prices: Bitcoin started what has
historically been a strong month about where it ended a dismal September,
holding over $19K.Insights: USDC's stablecoin-fueled model of money, in
which the dollar functions as an open 'protocol,' could allow innovation
to flourish. But healthy competition is a prerequisite.Catch the latest
episodes of CoinDesk TV for insightful interviews with crypto industry
leaders and analysis. And sign up for First Mover, our daily newsletter
putting the latest moves in crypto markets in context.
example_title: MLM 2
pipeline_tag: fill-mask
CryptoBERT is a pre-trained BERT (Bidirectional Encoder Representations from Transformers) model fine-tuned on a dataset of crypto-related news articles. It is designed to analyze and understand crypto news, providing valuable insights into the rapidly evolving world of cryptocurrencies.
Features
Domain-Specific Knowledge: Trained on a diverse dataset of crypto news, CryptoBERT captures domain-specific information, enabling it to understand the unique language and context of the cryptocurrency space.
Sentiment Analysis: CryptoBERT is capable of sentiment analysis, helping you gauge the overall sentiment expressed in crypto news articles, whether it's positive, negative, or neutral.
Named Entity Recognition (NER): The model excels in identifying key entities such as cryptocurrency names, organizations, and important figures, enhancing its ability to extract relevant information.
Fine-tuned for Crypto Jargon: CryptoBERT is fine-tuned to recognize and understand the specialized jargon commonly used in the crypto industry, ensuring accurate interpretation of news articles.