Repository files navigation

tqdm derives from the Arabic wordtaqaddum (تقدّم) which can mean "progress,"and is an abbreviation for "I love you so much" in Spanish (te quiero demasiado).

Instantly make your loops show a smart progress meter - just wrap anyiterable withtqdm(iterable), and you're done!

fromtqdmimporttqdmforiintqdm(range(10000)):    ...

76%|████████████████████████        | 7568/10000 [00:33<00:10, 229.00it/s]

trange(N) can be also used as a convenient shortcut fortqdm(range(N)).

It can also be executed as a module with pipes:

$ seq 9999999| tqdm --bytes| wc -l75.2MB [00:00, 217MB/s]9999999$ tar -zcf - docs/| tqdm --bytes --total`du -sb docs/| cut -f1` \> backup.tgz 32%|██████████▍| 8.89G/27.9G [00:42<01:31, 223MB/s]

Overhead is low -- about 60ns per iteration (80ns withtqdm.gui), and isunit tested against performance regression.By comparison, the well-establishedProgressBar hasan 800ns/iter overhead.

In addition to its low overhead,tqdm uses smart algorithms to predictthe remaining time and to skip unnecessary iteration displays, which allowsfor a negligible overhead in most cases.

tqdm works on any platform(Linux, Windows, Mac, FreeBSD, NetBSD, Solaris/SunOS),in any console or in a GUI, and is also friendly with IPython/Jupyter notebooks.

tqdm does not require any dependencies (not evencurses!), justPython and an environment supportingcarriage return \r andline feed \n control characters.

pip install tqdm

Pull and install pre-releasedevel branch:

pip install"git+https://github.com/tqdm/tqdm.git@devel#egg=tqdm"

conda install -c conda-forge tqdm

There are 3 channels to choose from:

snap install tqdm# implies --stable, i.e. latest tagged releasesnap install tqdm  --candidate# master branchsnap install tqdm  --edge# devel branch

Note thatsnap binaries are purely for CLI use (notimport-able), andautomatically set upbash tab-completion.

docker pull tqdm/tqdmdocker run -i --rm tqdm/tqdm --help

There are other (unofficial) places wheretqdm may be downloaded, particularly for CLI use:

The list of all changes is available either on GitHub's Releases:, on thewiki, or on thewebsite.

tqdm is very versatile and can be used in a number of ways.The three main ones are given below.

Wraptqdm() around any iterable:

fromtqdmimporttqdmfromtimeimportsleeptext=""forcharintqdm(["a","b","c","d"]):sleep(0.25)text=text+char

trange(i) is a special optimised instance oftqdm(range(i)):

fromtqdmimporttrangeforiintrange(100):sleep(0.01)

Instantiation outside of the loop allows for manual control overtqdm():

pbar=tqdm(["a","b","c","d"])forcharinpbar:sleep(0.25)pbar.set_description("Processing %s"%char)

Manual control oftqdm() updates using awith statement:

withtqdm(total=100)aspbar:foriinrange(10):sleep(0.1)pbar.update(10)

If the optional variabletotal (or an iterable withlen()) isprovided, predictive stats are displayed.

with is also optional (you can just assigntqdm() to a variable,but in this case don't forget todel orclose() at the end:

pbar=tqdm(total=100)foriinrange(10):sleep(0.1)pbar.update(10)pbar.close()

Perhaps the most wonderful use oftqdm is in a script or on the commandline. Simply insertingtqdm (orpython -m tqdm) between pipes will passthrough allstdin tostdout while printing progress tostderr.

The example below demonstrate counting the number of lines in all Python filesin the current directory, with timing information included.

$time find. -name'*.py' -type f -exec cat\{}\;| wc -l857365real    0m3.458suser    0m0.274ssys     0m3.325s$time find. -name'*.py' -type f -exec cat\{}\;| tqdm| wc -l857366it [00:03, 246471.31it/s]857365real    0m3.585suser    0m0.862ssys     0m3.358s

Note that the usual arguments fortqdm can also be specified.

$ find. -name'*.py' -type f -exec cat\{}\;|    tqdm --unit loc --unit_scale --total 857366>> /dev/null100%|█████████████████████████████████| 857K/857K [00:04<00:00, 246Kloc/s]

Backing up a large directory?

$ tar -zcf - docs/| tqdm --bytes --total`du -sb docs/| cut -f1` \> backup.tgz 44%|██████████████▊| 153M/352M [00:14<00:18, 11.0MB/s]

This can be beautified further:

$ BYTES=$(du -sb docs/| cut -f1)$ tar -cf - docs/ \| tqdm --bytes --total"$BYTES" --desc Processing| gzip \| tqdm --bytes --total"$BYTES" --desc Compressed --position 1 \>~/backup.tgzProcessing: 100%|██████████████████████| 352M/352M [00:14<00:00, 30.2MB/s]Compressed:  42%|█████████▎| 148M/352M [00:14<00:19, 10.9MB/s]

Or done on a file level using 7-zip:

$ 7z a -bd -r backup.7z docs/| grep Compressing \| tqdm --total$(find docs/ -type f| wc -l) --unit files \| grep -v Compressing100%|██████████████████████████▉| 15327/15327 [01:00<00:00, 712.96files/s]

Pre-existing CLI programs already outputting basic progress information willbenefit fromtqdm's--update and--update_to flags:

$ seq 3 0.1 5| tqdm --total 5 --update_to --null100%|████████████████████████████████████| 5.0/5 [00:00<00:00, 9673.21it/s]$ seq 10| tqdm --update --null# 1 + 2 + ... + 10 = 55 iterations55it [00:00, 90006.52it/s]

The most common issues relate to excessive output on multiple lines, insteadof a neat one-line progress bar.

• Consoles in general: require support for carriage return (CR,\r).
  • Some cloud logging consoles which don't support\r properly(cloudwatch,K8s) may benefit fromexport TQDM_POSITION=-1.
• Nested progress bars:
  • Consoles in general: require support for moving cursors up to theprevious line. For example,IDLE,ConEmu andPyCharm (alsohere,here, andhere)lack full support.
  • Windows: additionally may require the Python modulecoloramato ensure nested bars stay within their respective lines.
• Unicode:
  • Environments which report that they support unicode will have solid smoothprogressbars. The fallback is anascii-only bar.
  • Windows consoles often only partially support unicode and thusoften require explicit ascii=True(alsohere). This is due toeither normal-width unicode characters being incorrectly displayed as"wide", or some unicode characters not rendering.
• Wrapping generators:
  • Generator wrapper functions tend to hide the length of iterables.tqdm does not.
  • Replacetqdm(enumerate(...)) withenumerate(tqdm(...)) ortqdm(enumerate(x), total=len(x), ...).The same applies tonumpy.ndenumerate.
  • Replacetqdm(zip(a, b)) withzip(tqdm(a), b) or evenzip(tqdm(a), tqdm(b)).
  • The same applies toitertools.
  • Some useful convenience functions can be found undertqdm.contrib.
• No intermediate output in docker-compose:usedocker-compose run instead ofdocker-compose up andtty: true.
• Overriding defaults via environment variables:e.g. in CI/cloud jobs,export TQDM_MININTERVAL=5 to avoid log spam.This override logic is handled by thetqdm.utils.envwrap decorator(useful independent oftqdm).

If you come across any other difficulties, browse and file.

(Since 19 May 2016)

classtqdm():"""  Decorate an iterable object, returning an iterator which acts exactly  like the original iterable, but prints a dynamically updating  progressbar every time a value is requested.  """@envwrap("TQDM_")# override defaults via env varsdef__init__(self,iterable=None,desc=None,total=None,leave=True,file=None,ncols=None,mininterval=0.1,maxinterval=10.0,miniters=None,ascii=None,disable=False,unit='it',unit_scale=False,dynamic_ncols=False,smoothing=0.3,bar_format=None,initial=0,position=None,postfix=None,unit_divisor=1000,write_bytes=False,lock_args=None,nrows=None,colour=None,delay=0):

• iterable:iterable, optional

  Iterable to decorate with a progressbar.Leave blank to manually manage the updates.

• desc:str, optional

  Prefix for the progressbar.

• total:int or float, optional

  The number of expected iterations. If unspecified,len(iterable) is used if possible. If float("inf") or as a lastresort, only basic progress statistics are displayed(no ETA, no progressbar).Ifgui is True and this parameter needs subsequent updating,specify an initial arbitrary large positive number,e.g. 9e9.

• leave:bool, optional

  If [default: True], keeps all traces of the progressbarupon termination of iteration.IfNone, will leave only ifposition is0.

• file:io.TextIOWrapper orio.StringIO, optional

  Specifies where to output the progress messages(default: sys.stderr). Usesfile.write(str) andfile.flush()methods. For encoding, seewrite_bytes.

• ncols:int, optional

  The width of the entire output message. If specified,dynamically resizes the progressbar to stay within this bound.If unspecified, attempts to use environment width. Thefallback is a meter width of 10 and no limit for the counter andstatistics. If 0, will not print any meter (only stats).

• mininterval:float, optional

  Minimum progress display update interval [default: 0.1] seconds.

• maxinterval:float, optional

  Maximum progress display update interval [default: 10] seconds.Automatically adjustsminiters to correspond tominintervalafter long display update lag. Only works ifdynamic_minitersor monitor thread is enabled.

• miniters:int or float, optional

  Minimum progress display update interval, in iterations.If 0 anddynamic_miniters, will automatically adjust to equalmininterval (more CPU efficient, good for tight loops).If > 0, will skip display of specified number of iterations.Tweak this andmininterval to get very efficient loops.If your progress is erratic with both fast and slow iterations(network, skipping items, etc) you should set miniters=1.

• ascii:bool or str, optional

  If unspecified or False, use unicode (smooth blocks) to fillthe meter. The fallback is to use ASCII characters " 123456789#".

• disable:bool, optional

  Whether to disable the entire progressbar wrapper[default: False]. If set to None, disable on non-TTY.

• unit:str, optional

  String that will be used to define the unit of each iteration[default: it].

• unit_scale:bool or int or float, optional

  If 1 or True, the number of iterations will be reduced/scaledautomatically and a metric prefix following theInternational System of Units standard will be added(kilo, mega, etc.) [default: False]. If any other non-zeronumber, will scaletotal andn.

• dynamic_ncols:bool, optional

  If set, constantly altersncols andnrows to theenvironment (allowing for window resizes) [default: False].

• smoothing:float, optional

  Exponential moving average smoothing factor for speed estimates(ignored in GUI mode). Ranges from 0 (average speed) to 1(current/instantaneous speed) [default: 0.3].

• bar_format:str, optional

  Specify a custom bar string formatting. May impact performance.[default: '{l_bar}{bar}{r_bar}'], wherel_bar='{desc}: {percentage:3.0f}%|' andr_bar='| {n_fmt}/{total_fmt} [{elapsed}<{remaining}, ''{rate_fmt}{postfix}]'Possible vars: l_bar, bar, r_bar, n, n_fmt, total, total_fmt,percentage, elapsed, elapsed_s, ncols, nrows, desc, unit,rate, rate_fmt, rate_noinv, rate_noinv_fmt,rate_inv, rate_inv_fmt, postfix, unit_divisor,remaining, remaining_s, eta.Note that a trailing ": " is automatically removed after {desc}if the latter is empty.

• initial:int or float, optional

  The initial counter value. Useful when restarting a progressbar [default: 0]. If using float, consider specifying{n:.3f}or similar inbar_format, or specifyingunit_scale.

• position:int, optional

  Specify the line offset to print this bar (starting from 0)Automatic if unspecified.Useful to manage multiple bars at once (eg, from threads).

• postfix:dict or*, optional

  Specify additional stats to display at the end of the bar.Callsset_postfix(**postfix) if possible (dict).

• unit_divisor:float, optional

  [default: 1000], ignored unlessunit_scale is True.

• write_bytes:bool, optional

  Whether to write bytes. If (default: False) will write unicode.

• lock_args:tuple, optional

  Passed torefresh for intermediate output(initialisation, iterating, and updating).

• nrows:int, optional

  The screen height. If specified, hides nested bars outside thisbound. If unspecified, attempts to use environment height.The fallback is 20.

• colour:str, optional

  Bar colour (e.g. 'green', '#00ff00').

• delay:float, optional

  Don't display until [default: 0] seconds have elapsed.

• delim:chr, optional

  Delimiting character [default: 'n']. Use '0' for null.N.B.: on Windows systems, Python converts 'n' to 'rn'.

• buf_size:int, optional

  String buffer size in bytes [default: 256]used whendelim is specified.

• bytes:bool, optional

  If true, will count bytes, ignoredelim, and defaultunit_scale to True,unit_divisor to 1024, andunit to 'B'.

• tee:bool, optional

  If true, passesstdin to bothstderr andstdout.

• update:bool, optional

  If true, will treat input as newly elapsed iterations,i.e. numbers to pass toupdate(). Note that this is slow(~2e5 it/s) since every input must be decoded as a number.

• update_to:bool, optional

  If true, will treat input as total elapsed iterations,i.e. numbers to assign toself.n. Note that this is slow(~2e5 it/s) since every input must be decoded as a number.

• null:bool, optional

  If true, will discard input (no stdout).

• manpath:str, optional

  Directory in which to install tqdm man pages.

• comppath:str, optional

  Directory in which to place tqdm completion.

• log:str, optional

  CRITICAL|FATAL|ERROR|WARN(ING)|[default: 'INFO']|DEBUG|NOTSET.

• out : decorated iterator.

classtqdm():defupdate(self,n=1):"""      Manually update the progress bar, useful for streams      such as reading files.      E.g.:      >>> t = tqdm(total=filesize) # Initialise      >>> for current_buffer in stream:      ...    ...      ...    t.update(len(current_buffer))      >>> t.close()      The last line is highly recommended, but possibly not necessary if      ``t.update()`` will be called in such a way that ``filesize`` will be      exactly reached and printed.      Parameters      ----------      n  : int or float, optional          Increment to add to the internal counter of iterations          [default: 1]. If using float, consider specifying ``{n:.3f}``          or similar in ``bar_format``, or specifying ``unit_scale``.      Returns      -------      out  : bool or None          True if a ``display()`` was triggered.      """defclose(self):"""Cleanup and (if leave=False) close the progressbar."""defclear(self,nomove=False):"""Clear current bar display."""defrefresh(self):"""      Force refresh the display of this bar.      Parameters      ----------      nolock  : bool, optional          If ``True``, does not lock.          If [default: ``False``]: calls ``acquire()`` on internal lock.      lock_args  : tuple, optional          Passed to internal lock's ``acquire()``.          If specified, will only ``display()`` if ``acquire()`` returns ``True``.      """defunpause(self):"""Restart tqdm timer from last print time."""defreset(self,total=None):"""      Resets to 0 iterations for repeated use.      Consider combining with ``leave=True``.      Parameters      ----------      total  : int or float, optional. Total to use for the new bar.      """defset_description(self,desc=None,refresh=True):"""      Set/modify description of the progress bar.      Parameters      ----------      desc  : str, optional      refresh  : bool, optional          Forces refresh [default: True].      """defset_postfix(self,ordered_dict=None,refresh=True,**tqdm_kwargs):"""      Set/modify postfix (additional stats)      with automatic formatting based on datatype.      Parameters      ----------      ordered_dict  : dict or OrderedDict, optional      refresh  : bool, optional          Forces refresh [default: True].      kwargs  : dict, optional      """@classmethoddefwrite(cls,s,file=sys.stdout,end="\n"):"""Print a message via tqdm (without overlap with bars)."""@propertydefformat_dict(self):"""Public API for read-only member access."""defdisplay(self,msg=None,pos=None):"""      Use ``self.sp`` to display ``msg`` in the specified ``pos``.      Consider overloading this function when inheriting to use e.g.:      ``self.some_frontend(**self.format_dict)`` instead of ``self.sp``.      Parameters      ----------      msg  : str, optional. What to display (default: ``repr(self)``).      pos  : int, optional. Position to ``moveto``        (default: ``abs(self.pos)``).      """@classmethod@contextmanagerdefwrapattr(cls,stream,method,total=None,bytes=True,**tqdm_kwargs):"""      stream  : file-like object.      method  : str, "read" or "write". The result of ``read()`` and          the first argument of ``write()`` should have a ``len()``.      >>> with tqdm.wrapattr(file_obj, "read", total=file_obj.size) as fobj:      ...     while True:      ...         chunk = fobj.read(chunk_size)      ...         if not chunk:      ...             break      """@classmethoddefpandas(cls,*targs,**tqdm_kwargs):"""Registers the current `tqdm` class with `pandas`."""deftrange(*args,**tqdm_kwargs):"""Shortcut for `tqdm(range(*args), **tqdm_kwargs)`."""

deftqdm.contrib.tenumerate(iterable,start=0,total=None,tqdm_class=tqdm.auto.tqdm,**tqdm_kwargs):"""Equivalent of `numpy.ndenumerate` or builtin `enumerate`."""deftqdm.contrib.tzip(iter1,*iter2plus,**tqdm_kwargs):"""Equivalent of builtin `zip`."""deftqdm.contrib.tmap(function,*sequences,**tqdm_kwargs):"""Equivalent of builtin `map`."""

classtqdm.notebook.tqdm(tqdm.tqdm):"""IPython/Jupyter Notebook widget."""classtqdm.auto.tqdm(tqdm.tqdm):"""Automatically chooses beween `tqdm.notebook` and `tqdm.tqdm`."""classtqdm.asyncio.tqdm(tqdm.tqdm):"""Asynchronous version."""@classmethoddefas_completed(cls,fs,*,loop=None,timeout=None,total=None,**tqdm_kwargs):"""Wrapper for `asyncio.as_completed`."""classtqdm.gui.tqdm(tqdm.tqdm):"""Matplotlib GUI version."""classtqdm.tk.tqdm(tqdm.tqdm):"""Tkinter GUI version."""classtqdm.rich.tqdm(tqdm.tqdm):"""`rich.progress` version."""classtqdm.keras.TqdmCallback(keras.callbacks.Callback):"""Keras callback for epoch and batch progress."""classtqdm.dask.TqdmCallback(dask.callbacks.Callback):"""Dask callback for task progress."""

Thetqdm.contrib package also contains experimental modules:

• tqdm.contrib.itertools: Thin wrappers arounditertools
• tqdm.contrib.concurrent: Thin wrappers aroundconcurrent.futures
• tqdm.contrib.slack: Posts toSlack bots
• tqdm.contrib.discord: Posts toDiscord bots
• tqdm.contrib.telegram: Posts toTelegram bots
• tqdm.contrib.bells: Automagically enables all optional features
  • auto,pandas,slack,discord,telegram

• See theexamplesfolder;
• import the module and runhelp();
• consult thewiki;
  • this has anexcellent articleon how to make agreat progressbar;
• check out theslides from PyData London, or
• run the.

Custom information can be displayed and updated dynamically ontqdm barswith thedesc andpostfix arguments:

fromtqdmimporttqdm,trangefromrandomimportrandom,randintfromtimeimportsleepwithtrange(10)ast:foriint:# Description will be displayed on the leftt.set_description('GEN %i'%i)# Postfix will be displayed on the right,# formatted automatically based on argument's datatypet.set_postfix(loss=random(),gen=randint(1,999),str='h',lst=[1,2])sleep(0.1)withtqdm(total=10,bar_format="{postfix[0]} {postfix[1][value]:>8.2g}",postfix=["Batch", {"value":0}])ast:foriinrange(10):sleep(0.1)t.postfix[1]["value"]=i/2t.update()

Points to remember when using{postfix[...]} in thebar_format string:

• postfix also needs to be passed as an initial argument in a compatibleformat, and
• postfix will be auto-converted to a string if it is adict-likeobject. To prevent this behaviour, insert an extra item into the dictionarywhere the key is not a string.

Additionalbar_format parameters may also be defined by overridingformat_dict, and the bar itself may be modified usingascii:

fromtqdmimporttqdmclassTqdmExtraFormat(tqdm):"""Provides a `total_time` format parameter"""@propertydefformat_dict(self):d=super().format_dicttotal_time=d["elapsed"]* (d["total"]or0)/max(d["n"],1)d.update(total_time=self.format_interval(total_time)+" in total")returndforiinTqdmExtraFormat(range(9),ascii=" .oO0",bar_format="{total_time}: {percentage:.0f}%|{bar}{r_bar}"):ifi==4:break

00:00 in total: 44%|0000.     | 4/9 [00:00<00:00, 962.93it/s]

Note that{bar} also supports a format specifier[width][type].

• width
  • unspecified (default): automatic to fillncols
  • int >= 0: fixed width overridingncols logic
  • int < 0: subtract from the automatic default
• type
  • a: ascii (ascii=True override)
  • u: unicode (ascii=False override)
  • b: blank (ascii=" " override)

This means a fixed bar with right-justified text may be created by using:bar_format="{l_bar}{bar:10}|{bar:-10b}right-justified"

tqdm supports nested progress bars. Here's an example:

fromtqdm.autoimporttrangefromtimeimportsleepforiintrange(4,desc='1st loop'):forjintrange(5,desc='2nd loop'):forkintrange(50,desc='3rd loop',leave=False):sleep(0.01)

For manual control over positioning (e.g. for multi-processing use),you may specifyposition=n wheren=0 for the outermost bar,n=1 for the next, and so on.However, it's best to check iftqdm can work without manualpositionfirst.

fromtimeimportsleepfromtqdmimporttrange,tqdmfrommultiprocessingimportPool,RLock,freeze_supportL=list(range(9))defprogresser(n):interval=0.001/ (n+2)total=5000text=f"#{n}, est.{interval*total:<04.2}s"for_intrange(total,desc=text,position=n):sleep(interval)if__name__=='__main__':freeze_support()# for Windows supporttqdm.set_lock(RLock())# for managing output contentionp=Pool(initializer=tqdm.set_lock,initargs=(tqdm.get_lock(),))p.map(progresser,L)

Note that in Python 3,tqdm.write is thread-safe:

fromtimeimportsleepfromtqdmimporttqdm,trangefromconcurrent.futuresimportThreadPoolExecutorL=list(range(9))defprogresser(n):interval=0.001/ (n+2)total=5000text=f"#{n}, est.{interval*total:<04.2}s"for_intrange(total,desc=text):sleep(interval)ifn==6:tqdm.write("n == 6 completed.")tqdm.write("`tqdm.write()` is thread-safe in py3!")if__name__=='__main__':withThreadPoolExecutor()asp:p.map(progresser,L)

tqdm can easily support callbacks/hooks and manual updates.Here's an example withurllib:

``urllib.urlretrieve`` documentation

[...]

If present, the hook function will be called once

on establishment of the network connection and once after each block read

thereafter. The hook will be passed three arguments; a count of blocks

transferred so far, a block size in bytes, and the total size of the file.

[...]

importurllib,osfromtqdmimporttqdmurllib=getattr(urllib,'request',urllib)classTqdmUpTo(tqdm):"""Provides `update_to(n)` which uses `tqdm.update(delta_n)`."""defupdate_to(self,b=1,bsize=1,tsize=None):"""        b  : int, optional            Number of blocks transferred so far [default: 1].        bsize  : int, optional            Size of each block (in tqdm units) [default: 1].        tsize  : int, optional            Total size (in tqdm units). If [default: None] remains unchanged.        """iftsizeisnotNone:self.total=tsizereturnself.update(b*bsize-self.n)# also sets self.n = b * bsizeeg_link="https://caspersci.uk.to/matryoshka.zip"withTqdmUpTo(unit='B',unit_scale=True,unit_divisor=1024,miniters=1,desc=eg_link.split('/')[-1])ast:# all optional kwargsurllib.urlretrieve(eg_link,filename=os.devnull,reporthook=t.update_to,data=None)t.total=t.n

Inspired bytwine#242.Functional alternative inexamples/tqdm_wget.py.

It is recommend to useminiters=1 whenever there is potentiallylarge differences in iteration speed (e.g. downloading a file overa patchy connection).

Wrapping read/write methods

To measure throughput through a file-like object'sread orwritemethods, useCallbackIOWrapper:

fromtqdm.autoimporttqdmfromtqdm.utilsimportCallbackIOWrapperwithtqdm(total=file_obj.size,unit='B',unit_scale=True,unit_divisor=1024)ast:fobj=CallbackIOWrapper(t.update,file_obj,"read")whileTrue:chunk=fobj.read(chunk_size)ifnotchunk:breakt.reset()# ... continue to use `t` for something else

Alternatively, use the even simplerwrapattr convenience function,which would condense both theurllib andCallbackIOWrapper examplesdown to:

importurllib,osfromtqdmimporttqdmeg_link="https://caspersci.uk.to/matryoshka.zip"response=getattr(urllib,'request',urllib).urlopen(eg_link)withtqdm.wrapattr(open(os.devnull,"wb"),"write",miniters=1,desc=eg_link.split('/')[-1],total=getattr(response,'length',None))asfout:forchunkinresponse:fout.write(chunk)

Therequests equivalent is nearly identical:

importrequests,osfromtqdmimporttqdmeg_link="https://caspersci.uk.to/matryoshka.zip"response=requests.get(eg_link,stream=True)withtqdm.wrapattr(open(os.devnull,"wb"),"write",miniters=1,desc=eg_link.split('/')[-1],total=int(response.headers.get('content-length',0)))asfout:forchunkinresponse.iter_content(chunk_size=4096):fout.write(chunk)

Custom callback

tqdm is known for intelligently skipping unnecessary displays. To make acustom callback take advantage of this, simply use the return value ofupdate(). This is set toTrue if adisplay() was triggered.

fromtqdm.autoimporttqdmasstd_tqdmdefexternal_callback(*args,**kwargs):    ...classTqdmExt(std_tqdm):defupdate(self,n=1):displayed=super().update(n)ifdisplayed:external_callback(**self.format_dict)returndisplayed

Note thatbreak isn't currently caught by asynchronous iterators.This means thattqdm cannot clean up after itself in this case:

fromtqdm.asyncioimporttqdmasyncforiintqdm(range(9)):ifi==2:break

Instead, either callpbar.close() manually or use the context manager syntax:

fromtqdm.asyncioimporttqdmwithtqdm(range(9))aspbar:asyncforiinpbar:ifi==2:break

Due to popular demand we've added support forpandas -- here's an exampleforDataFrame.progress_apply andDataFrameGroupBy.progress_apply:

importpandasaspdimportnumpyasnpfromtqdmimporttqdmdf=pd.DataFrame(np.random.randint(0,100, (100000,6)))# Register `pandas.progress_apply` and `pandas.Series.map_apply` with `tqdm`# (can use `tqdm.gui.tqdm`, `tqdm.notebook.tqdm`, optional kwargs, etc.)tqdm.pandas(desc="my bar!")# Now you can use `progress_apply` instead of `apply`# and `progress_map` instead of `map`df.progress_apply(lambdax:x**2)# can also groupby:# df.groupby(0).progress_apply(lambda x: x**2)

In case you're interested in how this works (and how to modify it for yourown callbacks), see theexamplesfolder or import the module and runhelp().

Akeras callback is also available:

fromtqdm.kerasimportTqdmCallback...model.fit(...,verbose=0,callbacks=[TqdmCallback()])

Adask callback is also available:

fromtqdm.daskimportTqdmCallbackwithTqdmCallback(desc="compute"):    ...arr.compute()# or use callback globallycb=TqdmCallback(desc="global")cb.register()arr.compute()

IPython/Jupyter is supported via thetqdm.notebook submodule:

fromtqdm.notebookimporttrange,tqdmfromtimeimportsleepforiintrange(3,desc='1st loop'):forjintqdm(range(100),desc='2nd loop'):sleep(0.01)

In addition totqdm features, the submodule provides a native Jupyterwidget (compatible with IPython v1-v4 and Jupyter), fully working nested barsand colour hints (blue: normal, green: completed, red: error/interrupt,light blue: no ETA); as demonstrated below.

Thenotebook version supports percentage or pixels for overall width(e.g.:ncols='100%' orncols='480px').

It is also possible to lettqdm automatically choose betweenconsole or notebook versions by using theautonotebook submodule:

fromtqdm.autonotebookimporttqdmtqdm.pandas()

Note that this will issue aTqdmExperimentalWarning if run in a notebooksince it is not meant to be possible to distinguish betweenjupyter notebookandjupyter console. Useauto instead ofautonotebook to suppressthis warning.

Note that notebooks will display the bar in the cell where it was created.This may be a different cell from the one where it is used.If this is not desired, either

• delay the creation of the bar to the cell where it must be displayed, or
• create the bar withdisplay=False, and in a later cell calldisplay(bar.container):

fromtqdm.notebookimporttqdmpbar=tqdm(...,display=False)

# different celldisplay(pbar.container)

Thekeras callback has adisplay() method which can be used likewise:

fromtqdm.kerasimportTqdmCallbackcbk=TqdmCallback(display=False)

# different cellcbk.display()model.fit(...,verbose=0,callbacks=[cbk])

Another possibility is to have a single bar (near the top of the notebook)which is constantly re-used (usingreset() rather thanclose()).For this reason, the notebook version (unlike the CLI version) does notautomatically callclose() uponException.

fromtqdm.notebookimporttqdmpbar=tqdm()

# different celliterable=range(100)pbar.reset(total=len(iterable))# initialise with new `total`foriiniterable:pbar.update()pbar.refresh()# force print final status but don't `close()`

To change the default arguments (such as makingdynamic_ncols=True),simply use built-in Python magic:

fromfunctoolsimportpartialfromtqdmimporttqdmasstd_tqdmtqdm=partial(std_tqdm,dynamic_ncols=True)

For further customisation,tqdm may be inherited from to create custom callbacks (as with theTqdmUpTo exampleabove) or for custom frontends(e.g. GUIs such as notebook or plotting packages). In the latter case:

1. def __init__() to callsuper().__init__(..., gui=True) to disableterminalstatus_printer creation.
2. Redefine:close(),clear(),display().

Consider overloadingdisplay() to use e.g.self.frontend(**self.format_dict) instead ofself.sp(repr(self)).

Some submodule examples of inheritance:

• tqdm/notebook.py
• tqdm/gui.py
• tqdm/tk.py
• tqdm/contrib/slack.py
• tqdm/contrib/discord.py
• tqdm/contrib/telegram.py

You can use atqdm as a meter which is not monotonically increasing.This could be becausen decreases (e.g. a CPU usage monitor) ortotalchanges.

One example would be recursively searching for files. Thetotal is thenumber of objects found so far, whilen is the number of those objects whichare files (rather than folders):

fromtqdmimporttqdmimportos.pathdeffind_files_recursively(path,show_progress=True):files= []# total=1 assumes `path` is a filet=tqdm(total=1,unit="file",disable=notshow_progress)ifnotos.path.exists(path):raiseIOError("Cannot find:"+path)defappend_found_file(f):files.append(f)t.update()deflist_found_dir(path):"""returns os.listdir(path) assuming os.path.isdir(path)"""listing=os.listdir(path)# subtract 1 since a "file" we found was actually this directoryt.total+=len(listing)-1# fancy way to give info without forcing a refresht.set_postfix(dir=path[-10:],refresh=False)t.update(0)# may trigger a refreshreturnlistingdefrecursively_search(path):ifos.path.isdir(path):forfinlist_found_dir(path):recursively_search(os.path.join(path,f))else:append_found_file(path)recursively_search(path)t.set_postfix(dir=path)t.close()returnfiles

Usingupdate(0) is a handy way to lettqdm decide when to trigger adisplay refresh to avoid console spamming.

This is a work in progress (see#737).

Sincetqdm uses a simple printing mechanism to display progress bars,you should not write any message in the terminal usingprint() whilea progressbar is open.

To write messages in the terminal without any collision withtqdm bardisplay, a.write() method is provided:

fromtqdm.autoimporttqdm,trangefromtimeimportsleepbar=trange(10)foriinbar:# Print using tqdm class method .write()sleep(0.1)ifnot (i%3):tqdm.write("Done task %i"%i)# Can also use bar.write()

By default, this will print to standard outputsys.stdout. but you canspecify any file-like object using thefile argument. For example, thiscan be used to redirect the messages writing to a log file or class.

If using a library that can print messages to the console, editing the libraryby replacingprint() withtqdm.write() may not be desirable.In that case, redirectingsys.stdout totqdm.write() is an option.

To redirectsys.stdout, create a file-like class that will writeany input string totqdm.write(), and supply the argumentsfile=sys.stdout, dynamic_ncols=True.

A reusable canonical example is given below:

fromtimeimportsleepimportcontextlibimportsysfromtqdmimporttqdmfromtqdm.contribimportDummyTqdmFile@contextlib.contextmanagerdefstd_out_err_redirect_tqdm():orig_out_err=sys.stdout,sys.stderrtry:sys.stdout,sys.stderr=map(DummyTqdmFile,orig_out_err)yieldorig_out_err[0]# Relay exceptionsexceptExceptionasexc:raiseexc# Always restore sys.stdout/err if necessaryfinally:sys.stdout,sys.stderr=orig_out_errdefsome_fun(i):print("Fee, fi, fo,".split()[i])# Redirect stdout to tqdm.write() (don't forget the `as save_stdout`)withstd_out_err_redirect_tqdm()asorig_stdout:# tqdm needs the original stdout# and dynamic_ncols=True to autodetect console widthforiintqdm(range(3),file=orig_stdout,dynamic_ncols=True):sleep(.5)some_fun(i)# After the `with`, printing is restoredprint("Done!")

Similar tosys.stdout/sys.stderr as detailed above, consoleloggingmay also be redirected totqdm.write().

Warning: if also redirectingsys.stdout/sys.stderr, make sure toredirectlogging first if needed.

Helper methods are available intqdm.contrib.logging. For example:

importloggingfromtqdmimporttrangefromtqdm.contrib.loggingimportlogging_redirect_tqdmLOG=logging.getLogger(__name__)if__name__=='__main__':logging.basicConfig(level=logging.INFO)withlogging_redirect_tqdm():foriintrange(9):ifi==4:LOG.info("console logging redirected to `tqdm.write()`")# logging restored

tqdm implements a few tricks to increase efficiency and reduce overhead.

• Avoid unnecessary frequent bar refreshing:mininterval defines how longto wait between each refresh.tqdm always gets updated in the background,but it will display only everymininterval.
• Reduce number of calls to check system clock/time.
• mininterval is more intuitive to configure thanminiters.A clever adjustment systemdynamic_miniters will automatically adjustminiters to the amount of iterations that fit into timemininterval.Essentially,tqdm will check if it's time to print without actuallychecking time. This behaviour can be still be bypassed by manually settingminiters.

However, consider a case with a combination of fast and slow iterations.After a few fast iterations,dynamic_miniters will setminiters to alarge number. When iteration rate subsequently slows,miniters willremain large and thus reduce display update frequency. To address this:

• maxinterval defines the maximum time between display refreshes.A concurrent monitoring thread checks for overdue updates and forces onewhere necessary.

The monitoring thread should not have a noticeable overhead, and guaranteesupdates at least every 10 seconds by default.This value can be directly changed by setting themonitor_interval ofanytqdm instance (i.e.t = tqdm.tqdm(...); t.monitor_interval = 2).The monitor thread may be disabled application-wide by settingtqdm.tqdm.monitor_interval = 0 before instantiation of anytqdm bar.

You can buytqdm branded merch now!

All source code is hosted onGitHub.Contributions are welcome.

See theCONTRIBUTINGfile for more information.

Developers who have made significant contributions, ranked bySLoC(surviving lines of code,git fame-wMC --excl '\.(png|gif|jpg)$'),are:

Name	ID	SLoC	Notes
Casper da Costa-Luis	casperdcl	~80%	primary maintainer
Stephen Larroque	lrq3000	~9%	team member
Martin Zugnoni	martinzugnoni	~3%
Daniel Ecer	de-code	~2%
Richard Sheridan	richardsheridan	~1%
Guangshuo Chen	chengs	~1%
Helio Machado	0x2b3bfa0	~1%
Kyle Altendorf	altendky	<1%
Noam Yorav-Raphael	noamraph	<1%	original author
Matthew Stevens	mjstevens777	<1%
Hadrien Mary	hadim	<1%	team member
Mikhail Korobov	kmike	<1%	team member

A list is available onthis wiki page.

Open Source (OSI approved):

Citation information:

(Since 19 May 2016)