sizhky / torch_snippets

Utilities for simple needs

Home Page:https://sizhky.github.io/torch_snippets/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Utilities for simple needs

torch snippets does a lot of default importing for you

Whether it is numpy, pandas, matplotlib or the useful functions that are mentioned below Simply call

from torch_snippets import *

All the imports are lightweight and thus should not take more than a couple of seconds

Auxiliary Functions

There are simple functions that are overloaded to take inputs and perform repetitive tasks that usually take a few lines to write

Images

show, inspect, Glob, read, resize, rotate

Files and Paths

stem, Glob, parent, name, fname,

makedir, zip_files, unzip_file,

find, extn,

readlines, writelines

Lists

L, flatten

Dump and load python objects

loaddill,dumpdill

Misc

Tqdm, Timer, randint, Logger

Sets

unique, diff, choose, common

Pytorch Modules

Reshape and Permute (nn.Modules)

Report as Pytorch Lightning Callback

LightningReport

Charts

Chart from altair

and many more to come...

Install

pip install torch_snippets

Usage

%time from torch_snippets import *
CPU times: user 1.79 s, sys: 672 ms, total: 2.46 s
Wall time: 2.62 s
dir()
['AttrDict',
 'B',
 'BB',
 'Blank',
 'C',
 'Chart',
 'DataLoader',
 'Dataset',
 'Debug',
 'E',
 'Excep',
 'F',
 'Float',
 'Glob',
 'Image',
 'ImportEnum',
 'In',
 'Inf',
 'Info',
 'Int',
 'L',
 'LightningReport',
 'NullType',
 'Out',
 'P',
 'PIL',
 'Path',
 'Permute',
 'PrettyString',
 'Report',
 'Reshape',
 'Self',
 'ShowPrint',
 'Stateful',
 'Str',
 'StrEnum',
 'T',
 'Timer',
 'Tqdm',
 'Warn',
 '_',
 '__',
 '___',
 '__builtin__',
 '__builtins__',
 '__doc__',
 '__loader__',
 '__name__',
 '__package__',
 '__spec__',
 '_dh',
 '_i',
 '_i1',
 '_i2',
 '_ih',
 '_ii',
 '_iii',
 '_oh',
 'add',
 'add_props',
 'alt',
 'anno_ret',
 'annotations',
 'arg0',
 'arg1',
 'arg2',
 'arg3',
 'arg4',
 'argnames',
 'argwhere',
 'attrdict',
 'basic_repr',
 'bbfy',
 'bind',
 'camel2snake',
 'charts',
 'choose',
 'chunked',
 'class2attr',
 'common',
 'compose',
 'copy_func',
 'crop_from_bb',
 'custom_dir',
 'cv2',
 'cycle',
 'defaults',
 'detuplify',
 'device',
 'df2bbs',
 'diff',
 'display',
 'dumpdill',
 'enlarge_bbs',
 'eq',
 'even_mults',
 'exec_local',
 'exit',
 'extn',
 'fastcores',
 'fastuple',
 'filter_dict',
 'filter_ex',
 'filter_keys',
 'filter_values',
 'find',
 'first',
 'flatten',
 'fname',
 'fname2',
 'ge',
 'gen',
 'get_class',
 'get_ipython',
 'getattrs',
 'glob',
 'groupby',
 'gt',
 'hasattrs',
 'ifnone',
 'ignore_exceptions',
 'in_',
 'inspect',
 'instantiate',
 'inum_methods',
 'is_',
 'is_array',
 'is_not',
 'isdir',
 'jitter',
 'last_index',
 'le',
 'line',
 'lines',
 'listify',
 'load_torch_model_weights_to',
 'loaddill',
 'loader',
 'logger',
 'lt',
 'lzip',
 'makedir',
 'map_ex',
 'maps',
 'maybe_attr',
 'md5',
 'merge',
 'mk_class',
 'mul',
 'ne',
 'nested_attr',
 'nested_idx',
 'nn',
 'not_',
 'now',
 'np',
 'null',
 'num_cpus',
 'num_methods',
 'nunique',
 'optim',
 'os',
 'otherwise',
 'pad',
 'parent',
 'partialler',
 'patch',
 'patch_property',
 'patch_to',
 'pd',
 'pdb',
 'pdfilter',
 'pl',
 'plt',
 'properties',
 'puttext',
 'quit',
 'rand',
 'randint',
 'range_of',
 're',
 'read',
 'readPIL',
 'readlines',
 'rect',
 'remove_duplicates',
 'rename_batch',
 'renumerate',
 'replicate',
 'resize',
 'risinstance',
 'rnum_methods',
 'rotate',
 'save_torch_model_weights_from',
 'see',
 'set_logging_level',
 'setattrs',
 'setify',
 'show',
 'shrink_bbs',
 'snake2camel',
 'sorted_ex',
 'stem',
 'stems',
 'stop',
 'store_attr',
 'str_enum',
 'sub',
 'subplots',
 'sys',
 'th',
 'to_absolute',
 'to_relative',
 'tonull',
 'torch',
 'torch_loader',
 'torchvision',
 'tqdm',
 'trange',
 'transforms',
 'true',
 'truediv',
 'try_attrs',
 'tuplify',
 'type_hints',
 'typed',
 'uint',
 'unique',
 'uniqueify',
 'unzip_file',
 'using_attr',
 'val2idx',
 'with_cast',
 'wrap_class',
 'write',
 'writelines',
 'xywh2xyXY',
 'zip_cycle',
 'zip_files']

About

Utilities for simple needs

https://sizhky.github.io/torch_snippets/

License:Apache License 2.0


Languages

Language:Jupyter Notebook 95.7%Language:Python 2.2%Language:HTML 2.1%Language:Makefile 0.0%