/TelegramTradeMsgBacktestML

Backtest telegram mesaage from whole channel about trading(as in stocks or crypto) using Mahcine Learnig (Named Entity Recognition or Token Classification) for zerodha and angel broking or for NSE or any one else.

Primary LanguageJupyter NotebookMIT LicenseMIT

TelegramTradeMsgBacktestML

Backtest telegram mesaage from whole channel about trading(as in stocks or crypto) using Mahcine Learnig (Named Entity Recognition or Token Classification).

Test now on HuggingFace : Screenshot from 2022-10-01 11-50-07

:: Link to test this model ::

What this repository contains? :

  1. Label data using LabelStudio NER(Named Entity Recognition or Token Classification) tool. Screenshot from 2022-09-30 12-28-50 convert to Screenshot from 2022-09-30 18-59-14

  2. Convert LabelStudio CSV or JSON to HuggingFace-autoTrain dataset conversion script Screenshot from 2022-10-01 10-36-03

  3. Train NER model on Hugginface-autoTrain. Screenshot from 2022-10-01 10-38-24

  4. Use Hugginface-autoTrain model to predict labels on new data in LabelStudio using LabelStudio-ML-Backend. Screenshot from 2022-10-01 10-41-07 Screenshot from 2022-10-01 10-42-36 Screenshot from 2022-10-01 10-44-56

  5. Define python function to predict labels using Hugginface-autoTrain model. Screenshot from 2022-10-01 10-47-08 Screenshot from 2022-10-01 10-47-25

  6. Only label new data from newly predicted-labels-dataset that has falsified labels. Screenshot from 2022-09-30 22-47-23

  7. Backtest Truely labelled dataset against real historical data of the stock using zerodha kiteconnect and jugaad_trader. Screenshot from 2022-10-01 00-05-55

  8. Evaluate total gained percentage since inception summation-wise and compounded and plot. Screenshot from 2022-10-01 00-06-59

  9. Listen to telegram channel for new LIVE messages using telegram API for algotrading. Screenshot from 2022-10-01 00-09-29

  10. Serve the app as flask web API for web request and respond to it as labelled tokens. Screenshot from 2022-10-01 00-12-12

  11. Outperforming or underperforming results of the telegram channel tips against exchange index by percentage. Screenshot from 2022-10-01 11-16-27

Place a custom order on hjLabs.in : https://hjLabs.in

📫 How to reach me

                         

Checkout Cool GitHub Other Repositories:

Checkout Our Other Products:

Some Cool Arduino and ESP8266 (or NodeMCU) IoT projects:

Checkout Our Awesome 3D GrabCAD Models:

Our HuggingFace Models :

Our HuggingFace Datasets :

Awesome Youtube Videos :

My Quirky Blog :

Awesome Android Apps :

Checkout Cool GitLab Other Repositories: