fixed earlystopping callback

Author: graviraja

week_0_project_setup/train.py

@@ -1,6 +1,7 @@
 import torch
 import pytorch_lightning as pl
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 
 from data import DataModule
 from model import ColaModel
@@ -13,14 +14,17 @@ def main():
  checkpoint_callback = ModelCheckpoint(
  dirpath="./models", monitor="val_loss", mode="min"
  )
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
 
  trainer = pl.Trainer(
  default_root_dir="logs",
  gpus=(1 if torch.cuda.is_available() else 0),
- max_epochs=1,
+ max_epochs=5,
  fast_dev_run=False,
  logger=pl.loggers.TensorBoardLogger("logs/", name="cola", version=1),
- callbacks=[checkpoint_callback],
+ callbacks=[checkpoint_callback, early_stopping_callback],
  )
  trainer.fit(cola_model, cola_data)
 

week_1_wandb_logging/train.py

@@ -3,6 +3,7 @@
 import pandas as pd
 import pytorch_lightning as pl
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -47,11 +48,15 @@ def main():
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=1,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=10,
  deterministic=True,
  # limit_train_batches=0.25,

week_2_hydra_config/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -59,11 +60,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  limit_train_batches=cfg.training.limit_train_batches,

week_3_dvc/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_4_onnx/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_5_docker/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_6_github_actions/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_7_ecr/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_8_serverless/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,

week_9_monitoring/train.py

@@ -7,6 +7,7 @@
 import pytorch_lightning as pl
 from omegaconf.omegaconf import OmegaConf
 from pytorch_lightning.callbacks import ModelCheckpoint
+from pytorch_lightning.callbacks.early_stopping import EarlyStopping
 from pytorch_lightning.loggers import WandbLogger
 
 from data import DataModule
@@ -60,11 +61,15 @@ def main(cfg):
  mode="min",
  )
 
+ early_stopping_callback = EarlyStopping(
+ monitor="val_loss", patience=3, verbose=True, mode="min"
+ )
+
  wandb_logger = WandbLogger(project="MLOps Basics", entity="raviraja")
  trainer = pl.Trainer(
  max_epochs=cfg.training.max_epochs,
  logger=wandb_logger,
- callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data)],
+ callbacks=[checkpoint_callback, SamplesVisualisationLogger(cola_data), early_stopping_callback],
  log_every_n_steps=cfg.training.log_every_n_steps,
  deterministic=cfg.training.deterministic,
  # limit_train_batches=cfg.training.limit_train_batches,