slackker 1.2.2

Python package for monitoring your Model training status in real-time on slack & telegram.

Introducing slackker! :fire:

slackker is a python package for monitoring your ML model training status in real-time on Slack & Telegram. It can send you update for your ML model training progress and send final report with graphs when the training finishes. So now you don't have to sit in front of the machine all the time. You can quickly go and grab coffee :coffee: downstairs or run some errands and still keep tracking the progress while on the move without loosing your peace of mind.

Requirements: slack_sdk>=3.19.0 and matplotlib

Table of contents :notebook:

• Installation
• Getting started with slackker callbacks
  • Setup slackker
  • Use slackker callbacks with Keras
    • Import slackker for Keras
    • Create slackker object for keras
    • Call slackker object into model.fit()
    • Final code for Keras
  • Use slackker callbacks with Lightning
    • Import slackker for Lightning
    • Create slackker object for lightning
    • Call slackker object in Trainer module
    • Final code for Lightning
• Support
• Citation
• Maintainer

Installation :arrow_down:

• Install slackker from PyPi is recommended. slackker is compatible with Python >= 3.6 and runs on Linux, MacOS X and Windows.
• Installing slackker in your environment is easy. Just use below pip command:

pip install slackker

Getting started with slackker callbacks

Setup slackker

• Slack: How to setup slackker for your slack channel
• Telegram: How to setup slackker for Telegram

Use slackker callbacks with Keras

Import slackker for Keras

Import slackker.callbacks with following line:

from slackker.callbacks.keras import SlackUpdate # for slack

or

from slackker.callbacks.keras import TelegramUpdate # for telegram

Create slackker object for keras

Create slackker object.

# for Slack
slackker_object = SlackUpdate(token="xoxb-123234234235-123234234235-adedce74748c3844747aed",
    channel="A04AAB77ABC",
    ModelName='Keras_NN',
    export='png',
    SendPlot=True,
    verbose=0)

or

# for Telegram
slackker_update = TelegramUpdate(token="1234567890:AAAAA_A111BBBBBCCC2DD3eEe44f5GGGgGG",
    ModelName='Simple_NN',
    export='png',
    SendPlot=True,
    verbose=0)

• token: (string) Slack app/Telegram token
• channel: (string) Slack channel where you want to receive updates
• ModelName: (string) Name for your model. This same name will be used in future for title of the generated plots.
• export: (string) default "png": Format for plots to be exported. (supported formats: eps, jpeg, jpg, pdf, pgf, png, ps, raw, rgba, svg, svgz, tif, tiff)
• SendPlots: (Bool) default False: If set to True it will export history of model, both training and validation, save it in the format given in export argument and send graphs to slack channel when training ends. If set to False it will not send exported graphs to slack channel.
• verbose: (int) default 0: You can sent the verbose level up to 3.
  • verbose = 0 No logging
  • verbose = 1 Info logging
  • verbose = 2 Debug/In-depth logging

Call slackker object into model.fit()

Now you can call slackker object into callbacks argument just like any other callbacks object.

history = model.fit(x_train, 
                    y_train,
                    epochs = 3,
                    batch_size = 16,
                    verbose=1,
                    validation_data=(x_val,y_val),
                    callbacks=[slackker])

Final code for Keras

# Import library for keras
from slackker.callbacks.keras import slackUpdate

# Train-Test split for your keras model
x_train, x_test, y_train, y_test = train_test_split(x, y, train_size=0.8)
x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, train_size=0.8)

# Build keras model
model = Sequential()
model.add(Dense(8,activation='relu',input_shape = (IMG_WIDTH, IMG_HEIGHT, DEPTH)))
model.add(Dense(3,activation='softmax'))
model.compile(optimizer = 'rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])

# Create Slackker object
slack_update = slackUpdate(token="xoxb-123234234235-123234234235-adedce74748c3844747aed48499bb",
    channel="A04AAB77ABC",
    modelName='SampleModel',
    export='png',
    sendPlot=True,
    verbose=0)

# Call Slackker object in model.fit() callbacks
history = model.fit(x_train, 
                    y_train,
                    epochs = 3,
                    batch_size = 16,
                    verbose=1,
                    validation_data=(x_val,y_val),
                    callbacks=[slack_update])

Use slackker callbacks with Lightning

Import slackker for Lightning

Import slackker.callbacks with following line:

from slackker.callbacks.lightning import SlackUpdate # for slack

or

from slackker.callbacks.lightning import TelegramUpdate # for telegram

Create slackker object for lightning

# for Slack
slackker_update = SlackUpdate(token="xoxb-123234234235-123234234235-adedce74748c3844747aed",
    channel="A04AAB77ABC",
    ModelName='Lightning NN',
    TrackLogs=['train_loss', 'train_acc', 'val_loss', 'val_acc'],
    monitor="val_loss",
    export='png',
    SendPlot=True,
    verbose=0)

or

# for Telegram
slackker_update = TelegramUpdate(token="1234567890:AAAAA_A111BBBBBCCC2DD3eEe44f5GGGgGG",
    ModelName="Lightning NN Testing",
    TrackLogs=['train_loss', 'train_acc', 'val_loss', 'val_acc'],
    monitor="val_loss",
    export='png',
    SendPlot=True,
    verbose=0)

• token: (string) Slack app/Telegram token
• channel: (string) Slack channel where you want to receive updates
• ModelName: (string) Name for your model. This same name will be used in future for title of the generated plots.
• TrackLogs: (list) List of logs you want slackker to send.
• monitor: (string) This metric will be used to determine best Epoch
• export: (string) default "png": Format for plots to be exported. (supported formats: eps, jpeg, jpg, pdf, pgf, png, ps, raw, rgba, svg, svgz, tif, tiff)
• SendPlots: (Bool) default False: If set to True it will export history of model, both training and validation, save it in the format given in export argument and send graphs to slack channel when training ends. If set to False it will not send exported graphs to slack channel.
• verbose: (int) default 0: You can sent the verbose level up to 3.
  • verbose = 0 No logging
  • verbose = 1 Info logging
  • verbose = 2 Debug/In-depth logging

Call slackker object in Trainer module

Now you can call slackker object into callbacks argument just like any other callbacks object.

trainer = Trainer(max_epochs=2,callbacks=[slackker_update])

Final code for Lightning

import torch
import torch.nn as nn
from torch.utils.data import DataLoader
import torchvision as tv
import torch.nn.functional as F
from lightning.pytorch import LightningModule, Trainer
from lightning.pytorch.callbacks import ModelCheckpoint, Callback
from lightning.pytorch.loggers import CSVLogger

from slackker.callbacks.lightning import SlackUpdate
from slackker.callbacks.lightning import TelegramUpdate

class LightningModel(LightningModule):
    def __init__(self):
        super().__init__()
        self.fc1 = nn.Linear(28*28,256)
        self.fc2 = nn.Linear(256,128)
        self.out = nn.Linear(128,10)

    def forward(self, x):
        batch_size, _, _, _ = x.size()
        x = x.view(batch_size,-1)
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        return self.out(x)

    def configure_optimizers(self):
        optimizer = torch.optim.Adam(self.parameters(), lr=1e-3)
        return optimizer

    def training_step(self, batch, batch_idx):
        x, y = batch
        y_hat = self.forward(x)

        # calculate Loss
        loss = F.cross_entropy(y_hat,y)

        #calculate accuracy
        _, predictions = torch.max(y_hat, dim=1)
        correct_predictions = torch.sum(predictions == y)
        accuracy = correct_predictions / y.shape[0]

        self.log("train_loss", loss, on_epoch=True)
        self.log("train_acc", accuracy, on_epoch=True)

        return loss

    def validation_step(self, batch, batch_idx):
        x, y = batch
        y_hat = self.forward(x)

        # calculate Loss
        loss = F.cross_entropy(y_hat,y)

        #calculate accuracy
        _, predictions = torch.max(y_hat, dim=1)
        correct_predictions = torch.sum(predictions == y)
        accuracy = correct_predictions / y.shape[0]

        self.log("val_loss", loss)
        self.log("val_acc", accuracy)

        return loss

train_data = tv.datasets.MNIST(".", train=True, download=True, transform=tv.transforms.ToTensor())
test_data = tv.datasets.MNIST(".", train=False, download=True, transform=tv.transforms.ToTensor())
train_loader = DataLoader(train_data, batch_size=128)
test_loader = DataLoader(test_data, batch_size=128)

model = LightningModel()

# slackker checkpoint for slack
slackker_update = SlackUpdate(token="xoxb-123234234235-123234234235-adedce74748c3844747aed",
    channel="A04AAB77ABC",
    ModelName='Lightning NN',
    TrackLogs=['train_loss', 'train_acc', 'val_loss', 'val_acc'],
    monitor="val_loss",
    export='png',
    SendPlot=True,
    verbose=0)

trainer = Trainer(max_epochs=2, callbacks=[slackker_update])
trainer.fit(model, train_loader, test_loader)

Support :sparkles:

If you get stuck, we’re here to help. The following are the best ways to get assistance working through your issue:

• Use our Github Issue Tracker for reporting bugs or requesting features. Contribution are the best way to keep slackker amazing :muscle:
• If you want to contribute please refer Contributor's Guide for how to contribute in a helpful and collaborative way :innocent:

Citation :page_facing_up:

Please cite slackker in your publications if this is useful for your project/research. Here is an example BibTeX entry:

@misc{siddheshgunjal2023slackker,
  title={slackker},
  author={Siddhesh Gunjal},
  year={2023},
  howpublished={\url{https://github.com/siddheshgunjal/slackker}},
}

Maintainer :sunglasses: