williamFalcon

import pytorch_lightning as pl
from pl_bolts.models.self_supervised import SimCLR
from pl_bolts.datamodules import ImagenetDataModule
from pl_bolts.models.self_supervised.simclr.transforms import SimCLRTrainDataTransform, SimCLREvalDataTransform

# data
datamodule = ImagenetDataModule(image_size=196)

# transforms
(c, h, w) = datamodule.size()

williamFalcon

import torch.utils.data as tud
import torch
from typing import List
import random
import nlp


def prepare_dataset(tokenizer, split="train", max_length=120, num_datapoints=100_000):
    """Prepares WikiText-103 dataset"""
    wikitext = nlp.load_dataset("wikitext", "wikitext-103-v1")

williamFalcon

import os
import pytorch_lightning as pl
from pl_bolts.models.regression import LogisticRegression
from pl_bolts.datamodules import ImagenetDataModule

# use imagenet
imagenet = ImagenetDataModule(data_dir=os.environ['IMGNET_PATH'], meta_root=os.environ['META_ROOT'], image_size=224, num_workers=32)

# input size is channels x height x width
input_dim = 3 * 224 * 224

williamFalcon

import pytorch_lightning.metrics.functional as plm

pred = torch.tensor([0, 1, 2, 3])
target = torch.tensor([0, 1, 2, 2])

# many popular classification metrics and more
plm.accuracy(pred, target)
plm.auc(pred, target)
plm.auroc(pred, target)
plm.average_precision(pred, target)

williamFalcon

#!/bin/bash
#
# script to extract ImageNet dataset
# ILSVRC2012_img_train.tar (about 138 GB)
# ILSVRC2012_img_val.tar (about 6.3 GB)
# make sure ILSVRC2012_img_train.tar & ILSVRC2012_img_val.tar in your current directory
#
#  https://github.com/facebook/fb.resnet.torch/blob/master/INSTALL.md
# 
#  train/

williamFalcon

from pytorch_lightning.loggers import Tensorboard, NeptuneLogger

neptune = NeptuneLogger()
tensorboard = Tensorboard()

model = ...

trainer = Trainer(logger=[neptune, tensorboard])
trainer.fit(model

williamFalcon

train = DataLoader(...)
test = DataLoader(...)
val = DataLoader(...)

trainer = Trainer()
model = LightningModule()

trainer.fit(model, 
            train_dataloader=train,
            val_dataloader=val, 

williamFalcon

import pytorch_lightning as pl

class MyAPICallback(pl.Callback):

    def on_init_start(self, trainer):
        requests.post('model started')

    def on_init_end(self, trainer):
        

williamFalcon

import pytorch_lightning as pl

class MyLoggingCallback(pl.Callback):

    def on_init_start(self, trainer):
        trainer.logger.experiment.log_tensorboard_images(...)

    def on_init_end(self, trainer):
        trainer.logger.experiment.save_or_something(...)

williamFalcon

import pytorch_lightning as pl

class MyPrintingCallback(pl.Callback):

    def on_init_start(self, trainer):
        print('Starting to init trainer!')

    def on_init_end(self, trainer):
        print('trainer is init now')