rahulvigneswaran

/train/n02112137/n02112137_3351.JPEG 96
/train/n02112137/n02112137_14715.JPEG 96
/train/n02112137/n02112137_8996.JPEG 96
/train/n02112137/n02112137_8952.JPEG 96
/train/n02112137/n02112137_482.JPEG 96
/train/n02112137/n02112137_2023.JPEG 96
/train/n02112137/n02112137_10146.JPEG 96
/train/n02112137/n02112137_13505.JPEG 96
/train/n02112137/n02112137_6212.JPEG 96
/train/n02112137/n02112137_9200.JPEG 96

rahulvigneswaran

rahulvigneswaran

rahulvigneswaran

def fine_labels():
    fine_labels = [
        'apple',  # id 0
        'aquarium_fish',
        'baby',
        'bear',
        'beaver',
        'bed',
        'bee',
        'beetle',

rahulvigneswaran

# ========================================================================================================================
import openai
openai.api_key = 
# ========================================================================================================================

# >>> len(a[31:60]), len(a[60:])
# (29, 40)

import os
from tqdm import tqdm, trange

rahulvigneswaran

import torchvision

model_conv = torchvision.models.resnet18(pretrained=True)

for name, param in model_conv.named_parameters():
    if not("fc" in name):
      param.requires_grad = False
    print(name, param.requires_grad)

rahulvigneswaran

# From Polyfjord - https://youtu.be/Q9IRHmfGDrU?t=17

names = ["gold",
         "chilli_pepper",
         "crystal_blue",
         "aquamarine",
         "army_green",
         "neon_yellow",
         "caramel",
         "pumpkin_orange",

rahulvigneswaran

import pprint as pp

fine_labels = [
    'apple',  # id 0
    'aquarium_fish',
    'baby',
    'bear',
    'beaver',
    'bed',
    'bee',

rahulvigneswaran

Table of Contents

• CPU
• GPU
• Disk Space

CPU

• Dont use CPUs to train. This will slow experiments of others drastically. Almost always try to run your exps on GPUs excluding certain edge cases.

GPU

• Before you start a run on a GPU that someone else is already using, please do try and run first on an empty GPU, to check the GPU memory your experiment will occupy. If it exceeds the available memory (Use gpustat --watch to find the usage), PLEASE DONT RUN ON IT. This will crash the other person's runs.

rahulvigneswaran

#!/bin/bash
#
# script to download and 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
#