PredictorProjGPT

Introdução

A estrutura de prompt PredictorProjGPT tem a capacidade de aceitar resultados de testes de desempenho como uma entrada e fornecer uma saída que inclui indicadores-chave de desempenho (KPIs) mapeados para seus objetos de dataframe correspondentes.

Documentação OpenAI : https://platform.openai.com/docs/libraries/python-library

Pré-requisitos

1. Chave de API OpenAI
2. Pacotes OpenAI
3. Pacotes Pandas
4. Pacotes Streamlit

Instalando Pacotes

1. Instale pacotes OpenAI

 pip install openai

2. Instale Pandas

 pip install pandas

3. Instale o Streamlit

 pip install streamlit

Saída

Solicitação de relatório de teste de desempenho de entrada

 A performance test was conducted on 24th April 2020 on a Salesforce application. The following observation was 
recorded by a performance tester after conclusion of the test.
Configuration:
1. There are total 40 transactions involved. The wait time between each transaction is 5 seconds.
2. Total Pacing Time is configured as 180 seconds.
    
Summary:
1. Average Response Time for the entire duration of the test was 6.5 seconds.
2. CPU Utilization peaked 75% in the first hour of the steady state.
3. Memory Consumption of 88% was observed at the last hour of the steady state.
4. Entire duration of the test was 5 hours. Endurance Test was performed.
5. Observed 10% error on the entire duration of the test
6. Predict the outcome

Resultado final


Instalação

1. Clone o repositório no diretório do seu projeto

 git clone 'enter repository url here'

2. Construir imagem

 docker build -t predictor-gpt-app .

2. Execute o contêiner

 docker run -it predictor-gpt-app

Atualização - 18/03/2024

 1. Recent Changes in OpenAI commands - https://github.com/openai/openai-python

 2. OpenAI secret key is now stored as environment variable