Simple tools for logging and visualizing, loading and training

Overview

TNT

TNT is a library providing powerful dataloading, logging and visualization utilities for Python. It is closely integrated with PyTorch and is designed to enable rapid iteration with any model or training regimen.

travis Documentation Status

Installation

TNT can be installed with pip. To do so, run:

pip install torchnet

If you run into issues, make sure that Pytorch is installed first.

You can also install the latest verstion from master. Just run:

pip install git+https://github.com/pytorch/[email protected]

To update to the latest version from master:

pip install --upgrade git+https://github.com/pytorch/[email protected]

About

TNT (imported as torchnet) is a framework for PyTorch which provides a set of abstractions for PyTorch aiming at encouraging code re-use as well as encouraging modular programming. It provides powerful dataloading, logging, and visualization utilities.

The project was inspired by TorchNet, and legend says that it stood for “TorchNetTwo”. Since the deprecation of TorchNet TNT has developed on its own.

For example, TNT provides simple methods to record model preformance in the torchnet.meter module and to log them to Visdom (or in the future, TensorboardX) with the torchnet.logging module.

In the future, TNT will also provide strong support for multi-task learning and transfer learning applications. It currently supports joint training data loading through torchnet.utils.MultiTaskDataLoader.

Most of the modules support NumPy arrays as well as PyTorch tensors on input, and so could potentially be used with other frameworks.

Getting Started

See some of the examples in https://github.com/pytorch/examples. We would like to include some walkthroughs in the docs (contributions welcome!).

[LEGACY] Differences with lua version

What's been ported so far:

  • Datasets:
    • BatchDataset
    • ListDataset
    • ResampleDataset
    • ShuffleDataset
    • TensorDataset [new]
    • TransformDataset
  • Meters:
    • APMeter
    • mAPMeter
    • AverageValueMeter
    • AUCMeter
    • ClassErrorMeter
    • ConfusionMeter
    • MovingAverageValueMeter
    • MSEMeter
    • TimeMeter
  • Engines:
    • Engine
  • Logger
    • Logger
    • VisdomLogger
    • MeterLogger [new, easy to plot multi-meter via Visdom]

Any dataset can now be plugged into torch.utils.DataLoader, or called .parallel(num_workers=8) to utilize multiprocessing.

Comments
  • Cleanup PT-D imports

    Cleanup PT-D imports

    Summary: The flow logic around torch.dist imports results in large number of pyre errors; would be preferable to just raise on importing as opposed to silently fail to import bindings.

    Cons: Some percentage (MacOS?) of users may have notebooks that imports pytorch.distributed, although would think small, since any attempt to call parts of the library would just fail...

    fb: TODO: assuming ok, will remove the 10's-100's of unused pyre ignores no longer required.

    Differential Revision: D39842273

    cla signed fb-exported Reverted 
    opened by dstaay-fb 10
  • Add more datasets

    Add more datasets

    Added:

    • SplitDataset
    • ConcatDataset

    Made load an optional parameter in ListDataset so that it can be used as TableDataset.

    Added seed to resample in ShuffleDataset

    I don't understand why elem_list has to torch.LongTensor if tensor in ListDataset (see here). It seems pointless.

    Sasank.

    opened by chsasank 8
  • Update the train() API parameter to take a state

    Update the train() API parameter to take a state

    Summary: Allow the states to be set up outside of the train() function, provides better flexibility. However, the state becomes an opaque object that might be modified accidentally.

    Suggest, we should move dataloader out of the state. and make the state immutable.

    Reviewed By: ananthsub

    Differential Revision: D40233501

    cla signed fb-exported 
    opened by zzzwen 7
  • Fix tests for CI

    Fix tests for CI

    Summary: Seeing errors like this in CI on trunk:

        from torchtnt.tests.runner.utils import DummyPredictUnit, generate_random_dataloader
    E   ModuleNotFoundError: No module named 'torchtnt.tests'
    

    moving the test utils to resolve it

    Differential Revision: D38774677

    cla signed fb-exported 
    opened by ananthsub 6
  • I found it would be failed when initialize MeterLogger server with `http://` head.

    I found it would be failed when initialize MeterLogger server with `http://` head.

    i found it would be failed when initialize MeterLogger server with http:// head.

    >>> a = MeterLogger(server="http://localhost", port=8097, env='hello')
    >>> b = torch.Tensor([1])
    >>> a.updateLoss(b, 'loss')
    >>> a.resetMeter(mode='Train', iepoch=1)
    

    Error message here

    Exception in user code:
    ------------------------------------------------------------
    Traceback (most recent call last):
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connection.py", line 141, in _new_conn
        (self.host, self.port), self.timeout, **extra_kw)
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/util/connection.py", line 60, in create_connection
        for res in socket.getaddrinfo(host, port, family, socket.SOCK_STREAM):
      File "/usr/lib/python3.6/socket.py", line 745, in getaddrinfo
        for res in _socket.getaddrinfo(host, port, family, type, proto, flags):
    socket.gaierror: [Errno -2] Name or service not known
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connectionpool.py", line 601, in urlopen
        chunked=chunked)
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connectionpool.py", line 357, in _make_request
        conn.request(method, url, **httplib_request_kw)
      File "/usr/lib/python3.6/http/client.py", line 1239, in request
        self._send_request(method, url, body, headers, encode_chunked)
      File "/usr/lib/python3.6/http/client.py", line 1285, in _send_request
        self.endheaders(body, encode_chunked=encode_chunked)
      File "/usr/lib/python3.6/http/client.py", line 1234, in endheaders
        self._send_output(message_body, encode_chunked=encode_chunked)
      File "/usr/lib/python3.6/http/client.py", line 1026, in _send_output
        self.send(msg)
      File "/usr/lib/python3.6/http/client.py", line 964, in send
        self.connect()
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connection.py", line 166, in connect
        conn = self._new_conn()
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connection.py", line 150, in _new_conn
        self, "Failed to establish a new connection: %s" % e)
    urllib3.exceptions.NewConnectionError: <urllib3.connection.HTTPConnection object at 0x7f6cc08ce198>: Failed to establish a new connection: [Errno -2] Name or service not known
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "/home/txk/.local/lib/python3.6/site-packages/requests/adapters.py", line 440, in send
        timeout=timeout
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/connectionpool.py", line 639, in urlopen
        _stacktrace=sys.exc_info()[2])
      File "/home/txk/.local/lib/python3.6/site-packages/urllib3/util/retry.py", line 388, in increment
        raise MaxRetryError(_pool, url, error or ResponseError(cause))
    urllib3.exceptions.MaxRetryError: HTTPConnectionPool(host='http', port=80): Max retries exceeded with url: //localhost:8097/events (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f6cc08ce198>: Failed to establish a new connection: [Errno -2] Name or service not known',))
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "/home/txk/.local/lib/python3.6/site-packages/visdom/__init__.py", line 261, in _send
        data=json.dumps(msg),
      File "/home/txk/.local/lib/python3.6/site-packages/requests/api.py", line 112, in post
        return request('post', url, data=data, json=json, **kwargs)
      File "/home/txk/.local/lib/python3.6/site-packages/requests/api.py", line 58, in request
        return session.request(method=method, url=url, **kwargs)
      File "/home/txk/.local/lib/python3.6/site-packages/requests/sessions.py", line 508, in request
        resp = self.send(prep, **send_kwargs)
      File "/home/txk/.local/lib/python3.6/site-packages/requests/sessions.py", line 618, in send
        r = adapter.send(request, **kwargs)
      File "/home/txk/.local/lib/python3.6/site-packages/requests/adapters.py", line 508, in send
        raise ConnectionError(e, request=request)
    requests.exceptions.ConnectionError: HTTPConnectionPool(host='http', port=80): Max retries exceeded with url: //localhost:8097/events (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f6cc08ce198>: Failed to establish a new connection: [Errno -2] Name or service not known',))
    
    
    opened by TuXiaokang 6
  • Some code format error exists in `meterlogger.py`

    Some code format error exists in `meterlogger.py`

    error message like this

    >>> import torchnet
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "/usr/local/lib/python3.6/site-packages/torchnet/__init__.py", line 5, in <module>
        from . import logger
      File "/usr/local/lib/python3.6/site-packages/torchnet/logger/__init__.py", line 2, in <module>
        from .meterlogger import MeterLogger
      File "/usr/local/lib/python3.6/site-packages/torchnet/logger/meterlogger.py", line 16
        self.nclass = nclass
                           ^
    TabError: inconsistent use of tabs and spaces in indentation
    
    opened by TuXiaokang 6
  • Add Visdom logging to TNT

    Add Visdom logging to TNT

    Added the ability to log to Visdom (like Tensorboard, but better!). It is a port of this repo to TNT.

    screen shot 2017-08-03 at 3 27 10 pm

    This PR adds TNT support for the existing Visdom plots and includes an example that uses it for MNIST (above).

    opened by alexsax 6
  • how to use conda install torchnet?

    how to use conda install torchnet?

    In my ubuntu sever, pip is ok, but torchnet is only in 'pip list' and not in 'conda list', so when i run xx.py, error is no module named 'torchnet'. So i want to ask if i can use conda to install torchnet and how to install it? could anyone help me? thanks!

    opened by qilong-zhang 5
  • Replace loss[0] with loss.item() due to deprecation

    Replace loss[0] with loss.item() due to deprecation

    When using tnt's meterlogger a deprecation warning by PyTorch does appear:

    UserWarning: invalid index of a 0-dim tensor. This will be an error in PyTorch 0.
    5. Use tensor.item() to convert a 0-dim tensor to a Python number
    

    Issued by this line.

    Replacing loss[0] with loss.item() should fix that without any further consequences

    opened by sauercrowd 5
  • Generic tnt.Engine?

    Generic tnt.Engine?

    Shouldn't the tnt.engine.Engine be changed to tnt.engine.SGDEngine and tnt.engine.Engine be a generic engine?

    This is similar to torchnet which allowed extending engines to make meta-engines like train_val_engine.

    opened by karandwivedi42 5
  • Update LRScheduler instance checks

    Update LRScheduler instance checks

    Summary: After https://github.com/pytorch/pytorch/pull/88503 , these isinstance checks no longer pass. Example: https://github.com/pytorch/tnt/actions/runs/3434539858/jobs/5725945075

    Differential Revision: D41177335

    cla signed fb-exported 
    opened by ananthsub 4
  • Create torchtnt.utils.lr_scheduler.TLRScheduler

    Create torchtnt.utils.lr_scheduler.TLRScheduler

    Summary:

    Introduce torchtnt.utils.lr_scheduler.TLRScheduler to manage compatibility of torch with expose of LRScheduler https://github.com/pytorch/pytorch/pull/88503

    It seems that issue persist still with 1.13.1 (see #285). Replace the get_version logic with try/except. _LRscheduler.

    Fixes #285 https://github.com/pytorch/tnt/issues/285

    cla signed 
    opened by dmtrs 7
  • Issue with torch 1.13

    Issue with torch 1.13

    🐛 Describe the bug

    Can not import torchtnt.framework due to AttributeError.

    Traceback (most recent call last):
      File "/Users/foo/app/runner/trainer/__init__.py", line 4, in <module>
        import torchtnt.framework
      File "/Users/foo/Library/Caches/pypoetry/virtualenvs/sp-mcom8aNU-py3.9/lib/python3.9/site-packages/torchtnt/framework/__init__.py", line 7, in <module>
        from .auto_unit import AutoUnit
      File "/Users/foo/Library/Caches/pypoetry/virtualenvs/sp-mcom8aNU-py3.9/lib/python3.9/site-packages/torchtnt/framework/auto_unit.py", line 23, in <module>
        from torchtnt.framework.unit import EvalUnit, PredictUnit, TrainUnit
      File "/Users/foo/Library/Caches/pypoetry/virtualenvs/sp-mcom8aNU-py3.9/lib/python3.9/site-packages/torchtnt/framework/unit.py", line 24, in <module>
        TLRScheduler = torch.optim.lr_scheduler.LRScheduler
    AttributeError: module 'torch.optim.lr_scheduler' has no attribute 'LRScheduler'
    

    From torch documentation it does not seem to have a torch.optim.lr_scheduler in stable version. https://pytorch.org/docs/stable/optim.html. Please also see BC-breaking change after 1.10 that may affect optimizer and lr step.

    From requirements.txt it does not seem this is locked in particular torch version.

    What is current working supported torch version for torchtnt=0.0.4? Is there plans to have particular supported torch versions?

    Versions

    % poetry run python collect_env.py                                                                                                                                                                      ✹ ✭
    Collecting environment information...
    PyTorch version: 1.13.1
    Is debug build: False
    CUDA used to build PyTorch: None
    ROCM used to build PyTorch: N/A
    
    OS: macOS 10.14.6 (x86_64)
    GCC version: Could not collect
    Clang version: Could not collect
    CMake version: Could not collect
    Libc version: N/A
    
    Python version: 3.9.13 (main, Sep  3 2022, 22:36:35)  [Clang 10.0.1 (clang-1001.0.46.4)] (64-bit runtime)
    Python platform: macOS-10.14.6-x86_64-i386-64bit
    Is CUDA available: False
    CUDA runtime version: No CUDA
    CUDA_MODULE_LOADING set to: N/A
    GPU models and configuration: No CUDA
    Nvidia driver version: No CUDA
    cuDNN version: No CUDA
    HIP runtime version: N/A
    MIOpen runtime version: N/A
    Is XNNPACK available: True
    
    Versions of relevant libraries:
    [pip3] mypy-extensions==0.4.3
    [pip3] numpy==1.23.5
    [pip3] torch==1.13.1
    [pip3] torchtnt==0.0.4
    [conda] Could not collect
    
    opened by dmtrs 1
  • Support recreating dataloaders during loop

    Support recreating dataloaders during loop

    Summary: Sometimes users need to recreate the dataloaders at the start of the epoch during the overall training loop. To support this, we allow users to register a creation function when creating the state for training or fitting. We pass the state as an argument since users may want to use progress information such as the progress counters to reinitialize the dataloader.

    We don't add this support for evaluate or predict since those functions iterate through the corresponding dataloader just once.

    For fit, this allows flexibility to reload training & evaluation dataloaders independently during if desired

    Differential Revision: D40539580

    cla signed fb-exported 
    opened by ananthsub 3
Releases(0.0.5.1)
🏖 Keras Implementation of Painting outside the box

Keras implementation of Image OutPainting This is an implementation of Painting Outside the Box: Image Outpainting paper from Standford University. So

Bendang 1.1k Dec 10, 2022
Implementation of the state-of-the-art vision transformers with tensorflow

ViT Tensorflow This repository contains the tensorflow implementation of the state-of-the-art vision transformers (a category of computer vision model

Mohammadmahdi NouriBorji 2 Mar 16, 2022
DeepLab resnet v2 model in pytorch

pytorch-deeplab-resnet DeepLab resnet v2 model implementation in pytorch. The architecture of deepLab-ResNet has been replicated exactly as it is from

Isht Dwivedi 601 Dec 22, 2022
Code for Emergent Translation in Multi-Agent Communication

Emergent Translation in Multi-Agent Communication PyTorch implementation of the models described in the paper Emergent Translation in Multi-Agent Comm

Facebook Research 75 Jul 15, 2022
Blind Image Super-resolution with Elaborate Degradation Modeling on Noise and Kernel

Blind Image Super-resolution with Elaborate Degradation Modeling on Noise and Kernel This repository is the official PyTorch implementation of BSRDM w

Zongsheng Yue 69 Jan 05, 2023
Code for the KDD 2021 paper 'Filtration Curves for Graph Representation'

Filtration Curves for Graph Representation This repository provides the code from the KDD'21 paper Filtration Curves for Graph Representation. Depende

Machine Learning and Computational Biology Lab 16 Oct 16, 2022
IMBENS: class-imbalanced ensemble learning in Python.

IMBENS: class-imbalanced ensemble learning in Python. Links: [Documentation] [Gallery] [PyPI] [Changelog] [Source] [Download] [知乎/Zhihu] [中文README] [a

Zhining Liu 176 Jan 04, 2023
[NeurIPS 2021] Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods

Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods Large Scale Learning on Non-Homophilous Graphs: New Benchmark

60 Jan 03, 2023
Semi-supervised Video Deraining with Dynamical Rain Generator (CVPR, 2021, Pytorch)

S2VD Semi-supervised Video Deraining with Dynamical Rain Generator (CVPR, 2021) Requirements and Dependencies Ubuntu 16.04, cuda 10.0 Python 3.6.10, P

Zongsheng Yue 53 Nov 23, 2022
Fast, general, and tested differentiable structured prediction in PyTorch

Fast, general, and tested differentiable structured prediction in PyTorch

HNLP 1.1k Dec 16, 2022
Code for "Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance Consistency" paper

UNICORN 🦄 Webpage | Paper | BibTex PyTorch implementation of "Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance Consistency" pap

118 Jan 06, 2023
Neural networks applied in recognizing guitar chords using python, AutoML.NET with C# and .NET Core

Chord Recognition Demo application The demo application is written in C# with .NETCore. As of July 9, 2020, the only version available is for windows

Andres Mauricio Rondon Patiño 24 Oct 22, 2022
[NeurIPS2021] Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks

Exploring Architectural Ingredients of Adversarially Robust Deep Neural Networks Code for NeurIPS 2021 Paper "Exploring Architectural Ingredients of A

Hanxun Huang 26 Dec 01, 2022
Facebook AI Image Similarity Challenge: Descriptor Track

Facebook AI Image Similarity Challenge: Descriptor Track This repository contains the code for our solution to the Facebook AI Image Similarity Challe

Sergio MP 17 Dec 14, 2022
Code for paper "Learning to Reweight Examples for Robust Deep Learning"

learning-to-reweight-examples Code for paper Learning to Reweight Examples for Robust Deep Learning. [arxiv] Environment We tested the code on tensorf

Uber Research 261 Jan 01, 2023
Understanding Convolutional Neural Networks from Theoretical Perspective via Volterra Convolution

nnvolterra Run Code Compile first: make compile Run all codes: make all Test xconv: make npxconv_test MNIST dataset needs to be downloaded, converted

1 May 24, 2022
Scalable training for dense retrieval models.

Scalable implementation of dense retrieval. Training on cluster By default it trains locally: PYTHONPATH=.:$PYTHONPATH python dpr_scale/main.py traine

Facebook Research 90 Dec 28, 2022
A Python parser that takes the content of a text file and then reads it into variables.

Text-File-Parser A Python parser that takes the content of a text file and then reads into variables. Input.text File 1. What is your ***? 1. 18 -

Kelvin 0 Jul 26, 2021
Code release for BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images

BlockGAN Code release for BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images BlockGAN: Learning 3D Object-aware Scene Rep

41 May 18, 2022
SymmetryNet: Learning to Predict Reflectional and Rotational Symmetries of 3D Shapes from Single-View RGB-D Images

SymmetryNet SymmetryNet: Learning to Predict Reflectional and Rotational Symmetries of 3D Shapes from Single-View RGB-D Images ACM Transactions on Gra

26 Dec 05, 2022