https://github.com/amperecomputingai/ai-hackathon-2022
Public AI Hackathon content
https://github.com/amperecomputingai/ai-hackathon-2022
Last synced: 18 days ago
JSON representation
Public AI Hackathon content
- Host: GitHub
- URL: https://github.com/amperecomputingai/ai-hackathon-2022
- Owner: AmpereComputingAI
- Created: 2022-09-26T18:59:17.000Z (over 3 years ago)
- Default Branch: master
- Last Pushed: 2022-10-05T01:15:10.000Z (over 3 years ago)
- Last Synced: 2025-03-02T19:19:13.426Z (over 1 year ago)
- Language: Python
- Size: 6.84 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Ampere AI use cases
This is PyTorch based RoBERTa pre-trained model trained on tweeteval dataset which contains ~12,000 tweets for the sentiment classification tasks with three classes being positive, negative, neutral sentiments. TweetEval Dataset consists of seven heterogenous tasks in twitter, all framed as multi-class tweet classification.One of the used for this demo is Sentiment Analysis. The model can be downloaded from hugging face: https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest?text=Covid+cases+are+increasing+fast%21
In this Interactive demo, the user can input text and submit to get the corresponding sentiment class probabilities. This web app is developed using Gradio Python library.
# Steps to run the demo on the cloud instance
### SSH to cloud instance
> $ ssh -i \ ubuntu@\
### Launch the container
> $ wget https://raw.githubusercontent.com/AmpereComputingAI/ai-hackathon-2022/master/nlp/sentiment-analysis/start-docker-cloud-hackathon.sh
> $ bash start-docker-cloud-hackathon.sh
Starting sentiment-analysis-pt-aio-demo container
Docker container ID:173e054cd5a9
### Start the demo
> \# ./start-webapp.sh
>
> Getting Webapp URL ...........................
Webapp URL: http://\:7860/
### Open ports on the server
> Install *firewalld* package
\$ sudo apt purge ufw -y
\$ sudo apt install firewalld -y
\$ sudo firewall-cmd --permanent --zone=public --add-port=\<7860-7864>/tcp
\$ sudo firewall-cmd --reload
\$ sudo firewall-cmd --zone=public --list-ports
### View on the Browser
> Copy the URL and paste on the browser