https://github.com/qubvel/keras_telegram_callback

Telegram-bot callback for your Keras model
https://github.com/qubvel/keras_telegram_callback

callback keras telegram

Last synced: 15 days ago
JSON representation

Telegram-bot callback for your Keras model

• Host: GitHub
• URL: https://github.com/qubvel/keras_telegram_callback
• Owner: qubvel
• Created: 2018-04-23T20:44:12.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2019-04-27T18:23:45.000Z (over 5 years ago)
• Last Synced: 2024-10-03T12:42:08.482Z (about 1 month ago)
• Topics: callback, keras, telegram
• Language: Python
• Size: 6.84 KB
• Stars: 31
• Watchers: 3
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## Keras callback in Telegram

Telegram-bot callback for your Keras model  

This module allows your model to send learning history to your chat in Telegram messenger.  

### Configure your Telegram Callback

To start use this callback you need fo register bot in telegram and get your `telegram_id`. Follow instructions below.

#### Step 1. Register your telegram bot.  

 - Find `BotFather` in Telegram.

 - Follow `BotFather` instructions to register your bot in a few steps and get `token`.

#### Step 2. Get your ID.

 - Find `@my_id_bot` bot in Telegram and get your `telegram_id`. If bot is not working get `telegram_id` manually (step 2a).  

#### Step 2a. Get your ID (optional).

 - Find your bot in telegram and send message 'hello!'

 - Paste in your browser `api.telegram.org/bot/getUpdates` (use token you get in previous step), if you did everything correctly you will recive a JSON where you can find your `telegram_id`.

### Example

```python

from .callbacks import TelegramCallback

# load data, define and compile model

...

# create callback

config = {

    'token': '556983321:AAHO-bSWaIqcvHL91Xw12X18OWczFIpY1s0',   # paste your bot token

    'telegram_id': 123456789,                                   # paste your telegram_id

}

tg_callback = TelegramCallback(config)

# start training

model.fit(x, y, batch_size=32, callbacks=[tg_callback])

```

Congratulations! Now you will recive logs (losses, metrics, lr, etc.) in Telegram!