A natural language modeling framework based on PyTorch

Facebook Research

Last update: Dec 27, 2022

Related tags

Text Data & NLP pytext

Overview

PyText is a deep-learning based NLP modeling framework built on PyTorch. PyText addresses the often-conflicting requirements of enabling rapid experimentation and of serving models at scale. It achieves this by providing simple and extensible interfaces and abstractions for model components, and by using PyTorch’s capabilities of exporting models for inference via the optimized Caffe2 execution engine. We are using PyText in Facebook to iterate quickly on new modeling ideas and then seamlessly ship them at scale.

Core PyText features:

Production ready models for various NLP/NLU tasks:
- Text classifiers
  - Yoon Kim (2014): Convolutional Neural Networks for Sentence Classification
  - Lin et al. (2017): A Structured Self-attentive Sentence Embedding
- Sequence taggers
  - Lample et al. (2016): Neural Architectures for Named Entity Recognition
- Joint intent-slot model
  - Zhang et al. (2016): A Joint Model of Intent Determination and Slot Filling for Spoken Language Understanding
- Contextual intent-slot models
Distributed-training support built on the new C10d backend in PyTorch 1.0
Mixed precision training support through APEX (trains faster with less GPU memory on NVIDIA Tensor Cores)
Extensible components that allows easy creation of new models and tasks
Reference implementation and a pretrained model for the paper: Gupta et al. (2018): Semantic Parsing for Task Oriented Dialog using Hierarchical Representations
Ensemble training support

Installing PyText

PyText requires Python 3.6.1 or above.

To get started on a Cloud VM, check out our guide.

Get the source code:

  $ git clone https://github.com/facebookresearch/pytext
  $ cd pytext

Create a virtualenv and install PyText:

  $ python3 -m venv pytext_venv
  $ source pytext_venv/bin/activate
  (pytext_venv) $ pip install pytext-nlp

Detailed instructions and more installation options can be found in our Documentation. If you encounter issues with missing dependencies during installation, please refer to OS Dependencies.

Train your first text classifier

For this first example, we'll train a CNN-based text-classifier that classifies text utterances, using the examples in tests/data/train_data_tiny.tsv. The data and configs files can be obtained either by cloning the repository or by downloading the files manually from GitHub.

  (pytext_venv) $ pytext train < demo/configs/docnn.json

By default, the model is created in /tmp/model.pt

Now you can export your model as a caffe2 net:

  (pytext_venv) $ pytext export < demo/configs/docnn.json

You can use the exported caffe2 model to predict the class of raw utterances like this:

  (pytext_venv) $ pytext --config-file demo/configs/docnn.json predict <<< '{"text": "create an alarm for 1:30 pm"}'

More examples and tutorials can be found in Full Documentation.

Join the community

Facebook group: https://www.facebook.com/groups/pytext/

License

PyText is BSD-licensed, as found in the LICENSE file.

Comments

pytext: torch.quantization -> torch.ao.quantization
Summary: This changes the imports in the caffe2/torch/ao/nn to include the new import locations.

codemod -d pytext --extensions py 'torch.quantization' 'torch.ao.quantization'

Differential Revision: D31302214
CLA Signed Merged fb-exported
opened by z-a-f 18
Integration
Summary: First step integrate PET & Pytext, the idea is introduce a new trainer and by convert Pytext train_loop to a PET state generator. More things include in this diff:

create a unittest to make the E2E pass

create a flag to enable/disable PET in pytext

create util functions like intialize_coordinator

Differential Revision: D19806903
CLA Signed Merged fb-exported
opened by isunjin 12
Minor refactor and code rearrange on module.py

Summary: Minor refactor and code rearrange on module.py Goal is to reuse the Pytext embedding module methods for pytext module

Differential Revision: D25997360
CLA Signed Merged fb-exported

opened by mikekgfb 9
Convert matmuls to quantizable nn.Linear modules (#889)

Summary: Pull Request resolved: https://github.com/fairinternal/fairseq-py/pull/889

We are converting matmuls to quantizable nn.Linear modules in this diff. First let's test profile after the diff to see how low level operations are changing.

Reviewed By: jmp84, edunov, lly-zero-one, jhcross

Differential Revision: D17964796
CLA Signed Merged

opened by halilakin 9
Change model input tokens to optional

Summary: The Byte LSTM does not need the input of tokens, which it inherits from LSTM language model. Making it optional and pass it as None will allow to model to skip build vocab part and report confusing OOV problems.

Reviewed By: kmalik22

Differential Revision: D18253210
CLA Signed Merged

opened by FanW123 8
Id for all modules and torchscript compatibility workaround

Summary: We need a id for each module that was created for incremental decoding. Also this diff includes a workaround to make sure our models work with the new scripting api.

Differential Revision: D17636128
CLA Signed Merged

opened by arbabu123 8

TypeError: init() got an unexpected keyword argument 'dtype'

I followed tutorial of ATIS on https://pytext-pytext.readthedocs-hosted.com/en/latest/atis_tutorial.html. On my way to step 3: pytext train < sample_config.json this error showed up:

Traceback (most recent call last):
  File "/home/hiepph/miniconda3/bin/pytext", line 11, in <module>
    load_entry_point('pytext-nlp', 'console_scripts', 'pytext')()
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/core.py", line 722, in __call__
    return self.main(*args, **kwargs)
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/core.py", line 697, in main
    rv = self.invoke(ctx)
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/core.py", line 1066, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/core.py", line 895, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/core.py", line 535, in invoke
    return callback(*args, **kwargs)
  File "/home/hiepph/miniconda3/lib/python3.6/site-packages/click/decorators.py", line 17, in new_func
    return f(get_current_context(), *args, **kwargs)
  File "/home/hiepph/src/pytext/pytext/main.py", line 232, in train
    train_model(config)
  File "/home/hiepph/src/pytext/pytext/workflow.py", line 55, in train_model
    task = prepare_task(config, dist_init_url, device_id, rank, world_size)
  File "/home/hiepph/src/pytext/pytext/workflow.py", line 78, in prepare_task
    return create_task(config.task)
  File "/home/hiepph/src/pytext/pytext/task/task.py", line 38, in create_task
    return create_component(ComponentType.TASK, task_config, metadata, model_state)
  File "/home/hiepph/src/pytext/pytext/config/component.py", line 142, in create_component
    return cls.from_config(config, *args, **kwargs)
  File "/home/hiepph/src/pytext/pytext/task/task.py", line 81, in from_config
    featurizer=featurizer,
  File "/home/hiepph/src/pytext/pytext/config/component.py", line 147, in create_data_handler
    ComponentType.DATA_HANDLER, data_handler_config, *args, **kwargs
  File "/home/hiepph/src/pytext/pytext/config/component.py", line 142, in create_component
    return cls.from_config(config, *args, **kwargs)
  File "/home/hiepph/src/pytext/pytext/data/joint_data_handler.py", line 76, in from_config
    DatasetFieldName.DOC_WEIGHT_FIELD: FloatField(),
  File "/home/hiepph/src/pytext/pytext/fields/field.py", line 288, in __init__
    unk_token=None,
  File "/home/hiepph/src/pytext/pytext/fields/field.py", line 54, in __init__
    super().__init__(*args, **kwargs)
TypeError: __init__() got an unexpected keyword argument 'dtype'

This doesn't happen with Train your first model tutorial.

opened by hiepph 8

Optional Resorting in VocabBuilder.

Summary: Added the options to bypass resorting of vocabulary tokens in VocabBuilder when running add_to_file() and make_vocab()

Differential Revision: D28578974
CLA Signed Merged fb-exported

opened by I2304 7
Add support for hypothesis 5.x

Update to support hypothesis 5.x primarily by overriding the now-enforced default deadline of 200ms where appropriate.

Reviewed By: thatch

Differential Revision: D20323893
CLA Signed Merged fb-exported

opened by qwhelan 7
Add CompatibleTrainer, an adapter to use the Library with TaskTrainer
Summary:

we have some new components(datasets, transforms, models etc) created for PyText Lib

these components are generic and can be re-used in multiple trainers(e.g. lightning)

to reproduce the same results across Lightning trainer and old TaskTrainer, we need CompatibleTrainer as an adapter, which uses the new components with TaskTrainer and minimizes the duplication of the full train loop logic in TaskTrainer

Differential Revision: D21435152
CLA Signed Merged fb-exported
opened by hudeven 7
Dynamic Batch Scheduler Implementation
Summary: This diff adds support of dynamic batch training in pytext. It creates a new batcher that computes the batch size depending on the current epoch.

The diff implements two schedulers:

Linear: increases batch size linearly

Exponential: increases batch size exponentially

API

The dynamic batcher extends the pooling batcher so that most the arguments are there and are pretty consistent. It is important to note that dynamic batch sizes only affects the training batch size, not eval or test.

Dynamic batcher holds a new configuration object scheduler_config, this contains the information needed to compute dynamic batch sizes namely:

class SchedulerConfig(ModuleConfig): # the initial batch size used for training start_batch_size: int = 32 # the final or max batch size to use, any scheduler should # not go over this batch size end_batch_size: int = 256 # the number of epochs to increase the batch size over epoch_period: int = 10 # the batch size is kept constant for `step_size` number of epochs step_size: int = 1

Paper: https://arxiv.org/abs/1711.00489

Reviewed By: seayoung1112, ArmenAg

Differential Revision: D18900677
CLA Signed Merged fb-exported
opened by AkshatSh 7
CVE-2007-4559 Patch

Patching CVE-2007-4559

Hi, we are security researchers from the Advanced Research Center at Trellix. We have began a campaign to patch a widespread bug named CVE-2007-4559. CVE-2007-4559 is a 15 year old bug in the Python tarfile package. By using extract() or extractall() on a tarfile object without sanitizing input, a maliciously crafted .tar file could perform a directory path traversal attack. We found at least one unsantized extractall() in your codebase and are providing a patch for you via pull request. The patch essentially checks to see if all tarfile members will be extracted safely and throws an exception otherwise. We encourage you to use this patch or your own solution to secure against CVE-2007-4559. Further technical information about the vulnerability can be found in this blog.

If you have further questions you may contact us through this projects lead researcher Kasimir Schulz.

opened by TrellixVulnTeam 0

Question on AUC-PR Hinge Loss

In the implementation of "Scalable Learning of Non-Decomposable Objectives" at https://github.com/facebookresearch/pytext/blob/main/pytext/loss/loss.py ,

the positive weight is 1 + lambda * (1 - precision) instead of (1 + lambda) * (1 - precision). The old implementation in the tensorflow repo (gone now) was also 1 + lambda * (1 - precision)

To me (1 + lambda) looks like coming from the following equation in the paper (https://arxiv.org/pdf/1608.04802.pdf)

Dividing each side by N = |Y-| + |Y+| and multiplying (1-precision), we get the first term (1+lambda)(1-precision)(loss+)/N:

loss = (1+lambda)(loss+) + lambda ( precision/(1-precision) ) (loss-) - lambda (# positives)

per-sample loss = (1+lambda)(loss+)/N + lambda ( precision/(1-precision) ) (loss-)/N - lambda (# positives)/N

multiplied per-sample loss =
          (1+lambda)(1-precision)(loss+)/N + lambda * precision * (loss-)/N - lambda (1-precision) (# positives)/N
          ^^^^^^^^^^^^^^^^^^^^^^^

Why is it 1 + lambda * (1-precision) ?

        hinge_loss = loss_utils.weighted_hinge_loss(
            labels.unsqueeze(-1),
            logits.unsqueeze(-1) - self.biases,
            positive_weights=1.0 + lambdas * (1.0 - self.precision_values),
            negative_weights=lambdas * self.precision_values,
        )

opened by elbaro 0

docs: Fix a few typos
There are small typos in:

pytext/models/embeddings/int_weighted_multi_category_embedding.py

pytext/torchscript/batchutils.py

pytext/torchscript/module.py

pytext/torchscript/tokenizer/bpe.py

Fixes:

Should read dictionary rather than doctionary.

Should read prioritizing rather than proiritizing.

Should read input rather than intput.

Semi-automated pull request generated by https://github.com/timgates42/meticulous/blob/master/docs/NOTE.md
CLA Signed
opened by timgates42 0
ImportError: cannot import name 'metanet_pb2' from partially initialized module 'caffe2.proto'
Steps to reproduce

git clone https://github.com/facebookresearch/Clinical-Trial-Parser.git

install pytext by source https://pytext.readthedocs.io/en/master/installation.html#install-from-source

go build ./...

go test ./...

pytext train < src/resources/config/ner.json

Observed Results

from caffe2.proto import caffe2_pb2 File "/data/anaconda3/envs/pytext/lib/python3.8/site-packages/caffe2/proto/init.py", line 15, in from caffe2.proto import caffe2_pb2, metanet_pb2, torch_pb2 ImportError: cannot import name 'metanet_pb2' from partially initialized module 'caffe2.proto'

What happened? This could be a description, log output, etc.

Expected Results

I expect pytext to carry out the trainig

What did you expect to happen? completed training

Relevant Code

in https://github.com/facebookresearch/Clinical-Trial-Parser.git

// TODO(you): code here to reproduce the problem I have been trying to get this lib to work for three days. I have tried python 3.5, 3.6, 3.7, and 3.10. I have tried installing pytext using conda, pip, and build from source. googling ends up in dead trails where people continue to see the same problem, but previously raised tickets are closed. Please give indication how to get past this step?
opened by bhomass 1

No module named 'pytorch'

Steps to reproduce

Installed both torch and torchtext from the source code

Torch version: '1.9.0a0+git854cc53'
TorchText version: '0.11.0a0+05cb992'
Install pytext from source

cd /tmp 
git clone https://github.com/facebookresearch/pytext.git
cd pytext
sudo python3 setup.py install

Import pytext

import pytext

Observed Results

What happened? This could be a description, log output, etc.

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-1-0d876a9b1d3f> in <module>
----> 1 import pytext

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/__init__.py in <module>
    10 from caffe2.python import workspace
    11 from caffe2.python.predictor import predictor_exporter
---> 12 from pytext.data.sources.data_source import DataSource
    13 from pytext.task import load
    14 from pytext.task.new_task import NewTask

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/data/__init__.py in <module>
    10     NaturalBatchSampler,
    11 )
---> 12 from .data import Batcher, Data, PoolingBatcher, generator_iterator
    13 from .data_handler import BatchIterator, CommonMetadata, DataHandler
    14 from .disjoint_multitask_data import DisjointMultitaskData

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/data/data.py in <module>
    12 from pytext.utils.usage import log_class_usage
    13 
---> 14 from .sources import DataSource, RawExample, TSVDataSource
    15 from .sources.data_source import (
    16     GeneratorIterator,

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/data/sources/__init__.py in <module>
     2 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
     3 
----> 4 from .conllu import CoNLLUNERDataSource
     5 from .data_source import DataSource, RawExample
     6 from .dense_retrieval import DenseRetrievalDataSource

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/data/sources/conllu.py in <module>
     5 from typing import Dict, List, Optional, Type
     6 
----> 7 from pytext.data.sources.data_source import RootDataSource, SafeFileWrapper
     8 
     9 

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/data/sources/data_source.py in <module>
    10 from pytext.data.utils import shard
    11 from pytext.utils.data import Slot, parse_slot_string
---> 12 from pytext.utils.file_io import PathManager
    13 
    14 

/usr/local/lib/python3.6/dist-packages/pytext_nlp-0.3.3-py3.6.egg/pytext/utils/file_io.py in <module>
    10 # TODO: @stevenliu use PathManagerFactory after it's released to PyPI
    11 from iopath.common.file_io import HTTPURLHandler
---> 12 from pytorch.text.fb.utils import PATH_MANAGER as PathManager  # noqa
    13 
    14 

ModuleNotFoundError: No module named 'pytorch'

Expected Results

What did you expect to happen? No errors while importing pytext

Module causing this error pytorch.text.fb.utils is not mentioned in the requirements.txt. Is this an internal module to FB? Any recommendation to circumvent this error?

Thanks!

opened by anjali-chadha 6

Releases(v0.3.3)

v0.3.3(Jun 8, 2020)
New features

Add XLM-R document classification server + console (#1358)

MLP layer embed for float tensors and FloatListSeqTensorizer for List[List[[float]] features. (#1374)

Add class_accuracy in MultiLabelSoftClassificationMetrics (#1371)

Add an option to skip test run after models have been trained (#1372)

Support DP in PyText (#1366)

Support torchscriptify in multi_label_classification_layer (#1350)

Add custom metric class for reporting Joint model metrics (#1339)

MultiLabel-MultiClass Model for Joint Sequence Tagging (#1335)

Scripted tokenizer support for DocModel (#1314)

Bugfixes

Fixed metric reporter aggregation and output layer for the multi-label classification

Remove move_state_dict_to_gpu, which is causing CUDA OOM (#1367)

Fix Flow's default conversion of dict to AttrDict

Fix bug in ClassificationOutputLayer that pad_idx is never respected (#1347)

Serializing/Deserializing type Any: bugfix and simplification (#1344)

Fix RoBERTa Q&A Training Bug with multiple BoS tokens. (#1343)

Other

Better error message for misconfigured data fields

Replace deprecated integer division with floor division operator

Add informative prints to assert statements (#1360)

TorchScript: Put dense tensor on the same device with other input tensors (#1361)

Update PyTorch + ONNX (#1340)

Update PyTorch + ONNX (#1340)- binary ONNX

Update PR Template (#1349)

Reduce memory request for pytext train operator

Add 'contrib' directory for experimental code (#1333)

Source code(tar.gz)
Source code(zip)
pytext-nlp-0.3.3.tar.gz(344.88 KB)
pytext_nlp-0.3.3-py3-none-any.whl(491.60 KB)
v0.3.2(Apr 27, 2020)
New features

Add Roberta model into BertPairwiseModel (#1336)

Support read file from http URL (#1317)

add a new PyText get_num_examples_from_batch function in model (#1319)

Add support for length label smoothing (#1308)

Add new metrics type for Masked Seq2Seq Joint Model (#1304)

Add mask generator and strategy (#1302)

Add separate logging for label loss and length loss (#1294)

Add tensorizer support for masking of target tokens (#1297)

Add length prediction and basic masked generator (#1290)

Add self attention option to conv_encoder and conv_decoder (#1291)

Entity Saliency modeling on PyText: EntitySalienceMetricReporter/EntitySalienceTask

In-batch negative training for BertPairwiseModel

Support embedding from decoder (#1284)

Add dense features to Roberta

Add projection layer to HuggingFace encoder (#1273)

add PyText Embedding TorchScript Wrapper

Add option to pad missing label in LabelListTensorizer (#1269)

Integrate PET and Introduce ElasticTrainer (#1266)

support PoolingType in DocNN. (#1259)

Added WordSeqEmbedding (#1255)

Open source Assistant NLU seq2seq model (#1236)

Support multi label classification

BART in decoupled model

Bug fixes

Fix Incorrect State Dict Assumption (#1326)

Bug fix for "RoBERTaTensorizer object has no attribute is_input" (#1334)

Cast model output to cpu (#1329)

Fix OSS predict-py API (#1320)

Fix "calling median on empty tensor" issue in MR (#1322)

Add ScriptDoNothingTokenizer so that torchscriptification of SPM does not fail (#1316)

Fix creating generator everytime (#1301)

fix dense feature for fp16

Avoid edge cases with quantization by setting a known seed (#1295)

Make torchscript predictions even on empty text / token inputs

fix dense feature TorchScript typing (#1281)

avoid zero division error in metrics reporter (#1271)

Fix contiguous issue in bilstm export (#1270)

fix debug file generation for multilabel classification (#1247)

Fix fp16 optimizer attribute name

Other

Simplify contextual embedding dimension computation in PyText (#1331)

New Debug File for masked seq2seq

Move MockConfigLoader to OSS (#1324)

Pass in optimizer config instead of create_optimizer to trainer

Remove unnecessary torch.no_grad() block (#1323)

Fix Memory Issues in Metric Reporter for Classification Tasks over large Label Spaces

Add contextual embedding support to OS seq2seq model (#1299)

recover xlm_r tutorial notebook (#1305)

Enable controlling bias in MLP decoder

Migrate serving tutorial to TorchScript (#1310)

delete caffe2 export (#1307)

add whitelist for ONNX export

Use dynamic quantization api for BeamSearch (#1303)

Remove requirement that eos/bos be supplied for sequence export. (#1300)

Multicolumn support

Multicolumn support in torchscriptify

Add caching support to RawExample and batch predict API (#1298)

Add save-pytext-snapshot command to PyText cmdline (#1285)

Update with Whatsapp calling data + support dictionary features (#1293)

add arrange_caffe2_model_inputs in BaseModel (#1292)

Replace unit-tests on LMModel and FLLanguageModelingTask by LiteLMModel and FLLiteLMTask (#1296)

changes to make mbart work (#1911)

handle encoder and decoder embedding

Add tutorial for semantic parsing. (#1288)

Add new fb beam search with fused operator (#1287)

Move generator builder to constructor so that it can easily overridden. (#1286)

Torchscriptify ELTensorizer (#1282)

Torchscript export for Seq2Seq model (#1265)

Change Seq2Seq model from_config() to a more general api (#1280)

add max_seq_len to DocNN TorchScript model (#1279)

support XLM-R model Embedding in TorchScript (#1278)

Generic PyText Checkpoint Manager Interface (#1267)

Fix backward compatibility issue of pad_missing in LabelListTensorizer (#1277)

Update mean reduction in NLLLoss (#1272)

migrate pages.integrity.scam.docnn_models.xxx (#1275)

Unify model input for ByteTokensDocumentModel (#1274)

Torchscriptify TokenTensorizer

Allow dictionaries to overwrite entries with #fairseq:overwrite comment (#1073)

Make WordSeqEmbedding ONNX compatible

If the snapshot path provided is not valid, throw error (#1268)

support vocab filter by min count

Unify input for TorchScript Tensorizers and Models (#1256)

Torchscriptify XLM-R

Add class logging to task (#1264)

Add usage logging to exporter (#1262)

Add usage logging across models (#1263)

Usage logging on data classes (#1261)

GPT2 BPE add lower casing support (#1260)

FAISS Embedding Search Space [3/5]

Return len of tokens of each sequence in SeqTokenTensorizer (#1254)

Vocab Limited Pretrained Embedding [2/5] (#1248)

add Stage.OTHERS and allow TB to print to a seperate prefix not in (TRAIN, TEST, EVAL) (#1258)

Add option to skip 2 stage tokenizer and bpe decode sequences in the debug file (#1257)

Add Testcase for Wordpiece Tokenizer (#1249)

modify accuracy calculation for multi-label classification (#1244)

Enable tests in pytext/config:pytext_all_config_test

Introduce Class Usage Logging (#1243)

Make PyText compatible with Any type (#1242)

Make dict_embedding Torchscript friendly (#1240)

Support MultipleData for export and kd generation

delete flaky/broken tests (#1238)

Add support for returning start & end indices.

Source code(tar.gz)
Source code(zip)
pytext-nlp-0.3.2.tar.gz(339.35 KB)
pytext_nlp-0.3.2-py3-none-any.whl(482.20 KB)
0.3.1(Jan 15, 2020)
New features

Implement SquadQA tensorizer in TorchScript (#1211)

Add session data source for df (#1202)

Dynamic Batch Scheduler Implementation (#1200)

Implement loss aware sparsifier (#1204)

Ability to Fine-tune XLM-R for NER on CoNLL Datasets (#1201)

TorchScriptify Tokenizer after training (#1191)

Linear Layer only blockwise sparsifier (#478)

Adding performance graph to pytext models (#1192)

Enable inference on GPUs by moving tensors to specified device (#472)

Add support for learning from soft labels for Squad (MRC) models (#1188)

Create byte-aware model that can make byte predictions (#468)

Minimum Trust Lamb (#1186)

Allow model to take byte-level input and make byte-level prediction (#1187)

Scheduler with Warmup (#1184)

Implement LAMB optimizer (#1183)

CyclicLRScheduler (#1157)

PyText Entity Linking: ELTask and ELMetricReporter (#1165)

Bug fixes

Don't upgrade if Tensorizer already given (#504)

avoid torchscriptify on a ScriptModule (#1214)

Make tensorboard robust to NaN and Inf in model params (#1206)

Fix circleCLI Test broken in D19027834 (#1205)

Fix small bug in pytext vocabulary (#401)

Fix CircleCI failure caused by black and regex (#1199)

Fix CircleCI (#1194)

Fix Circle CI Test broken by D18880705 (#1190)

fix weight load for new fairseq checkpoints (#1189)

Fix Heirarchical intent and slot filling demo is broken (#1012) (#1151)

Fix index error in dict embedding when exported to Caffe2 (#1182)

Fix zero loss tensor in SquadOutputLayer (#1181)

qa fix for ignore_impossible=False

Other

Printing out error's underlying reason (#1227)

tidy file path in help text for invocation of docnn.json example (#1221)

PyText option to disable CUDA when testing. (#1223)

make augmented lstm compatible w other lstms (#1224)

Vocab recursive lookup (#1222)

Fix simple typo: valus -> value (#1219)

support using RoundRobin ProcessGroup in Distributed training (#1213)

Use PathManager for all I/O (#1198)

Make PathManager robust to API changes in fvcore (#1196)

Support for TVM training (BERT) (#1210)

Exit LM task if special token exists in text for ByteTensorizer (#1207)

Config adapter for pytext XLM (#1172)

Use TensorizerImpl for both training and inference for BERT, RoBERTa and XLM tensorizer (#1195)

Replace gluster paths with local file paths for NLG configs (#1197)

Make BERT Classification compatible with TSEs that return Encoded Layers.

implement BertTensorizerImpl and XLMTensorizerImpl (#1193)

Make is_input field of tensorizer configurable (#474)

BERTTensorizerBaseImpl to reimplement BERTTensorizerBase to be TorchScriptable (#1163)

Improve LogitsWorkflow to handle dumping of raw inputs and multiple output tensors (#683)

Accumulative blockwise pruning (#1170)

Patch for UnicodeDecodeError due to BPE. (#1179)

Add pre-loaded task as parameter to caffe2 batch prediction API

Specify CFLAGS to install fairseq in MacOS (#1175)

Resolve dependency conflict by specifying python-dateutil==2.8.0 (#1176)

Proper training behavior if setting do_eval=False (#1155)

Make DeepCNNRepresentation torchscriptable (#453)

Source code(tar.gz)
Source code(zip)
pytext-nlp-0.3.1.tar.gz(303.89 KB)
pytext_nlp-0.3.1-py3-none-any.whl(427.40 KB)
0.3.0(Nov 28, 2019)
New Features

RoBERTa and XLM-R

Integrate XLM-R into PyText (#1120)

Consolidate BERT, XLM and RobERTa Tensorizers (#1119)

Add XLM-R for joint model (#1135)

Open source Roberta (#1032)

Simple Transformer module components for RoBERTa (#1043)

RoBERTa models for document classification (#933)

Enable MLM training for RobertaEncoder (#1126)

Standardize RoBERTa Tensorizer Vocab Creation (#1113)

Make RoBERTa usable in more tasks including QA (#1017)

RoBERTa-QA JIT (#1088)

Unify GPT2BPE Tokenizer (#1110)

Adding Google SentencePiece as a Tokenizer (#1106)

TorchScript support

General torchscript module (#1134)

Support torchscriptify XLM-R (#1138)

Add support for torchscriptification of XLM intent slot models (#1167)

Script xlm tensorizer (#1118)

Refactor ScriptTensorizer with general tensorize API (#1117)

ScriptXLMTensorizer (#1123)

Add support for Torchscript export of IntentSlotOutputLayer and CRF (#1146)

Refactor ScriptTensorizor to support both text and tokens input (#1096)

Add torchscriptify API in tokenizer and tensorizer (#1055)

Add more stats in torchscript latency script (#1044)

Exported Roberta torchscript model include both traced_model and pre-processing logic (#1013)

Native Torchscript Wordpiece Tokenizer Op for BERTSquadQA, Torchscriptify BertSQUADQAModel (#879)

TorchScript-ify BERT training (#887)

Modify Return Signature of TorchScript BERT (#1058)

Implement BertTensorizer and RoBERTaTensorizer in TorchScript (#1053)

Others

FairseqModelEnsemble class (#1116)

Inverse Sqrt Scheduler (#1150)

Lazy modules (#1039)

Adopt Fairseq MemoryEfficientFP16Optimizer in PyText (#910)

Add RAdam (#952)

Add AdamW (#945)

Unify FP16&FP32 API (#1006)

Add precision at recall metric (#1079)

Added PandasDataSource (#1098)

Support testing Caffe2 model (#1097)

Add contextual feature support to export for Seq2Seq models

Convert matmuls to quantizable nn.Linear modules (#1304)

PyTorch eager mode implementation (#1072)

Implement Blockwise Sparsification (#1050)

Support Fairseq FP16Optimizer (#1008)

Make FP16OptimizerApex wrapper on Apex/amp (#1007)

Remove vocab from cuda (#955)

Add dense input to XLMModel (#997)

Replace tensorboardX with torch.utils.tensorboard (#1003)

Add mentioning of mixed precision training support (#643)

Sparsification for CRF transition matrix (#982)

Add dense feature normalization to Char-LSTM TorchScript model. (#986)

Cosine similarity support for BERT pairwise model training (#967)

Combine training data from multiple sources (#953)

Support visualization of word embeddings in Tensorboard (#969)

Decouple decoder and output layer creation in BasePairwiseModel (#973)

Drop rows with insufficient columns in TSV data source (#954)

Add use_config_from_snapshot option(load config from snapshot or current task) (#970)

Add predict function for NewTask (#936)

Use create_module to create CharacterEmbedding (#920)

Add XLM based joint model

Add ConsistentXLMModel (#913)

Optimize Gelu module for caffe2 export (#918)

Save best model's sub-modules when enabled (#912)

Documentation / Usability

XLM-R tutorial in notebook (#1159)

Update XLM-R OSS tutorial and add Google Colab link (#1168)

Update "raw_text" to "text" in tutorial (#1010)

Make tutorial more trivial (add git clone) (#1037)

Changes to make tutorial code simpler (#1002)

Fix datasource tutorial example (#998)

Handle long documents in squad qa datasource and models (#975)

Fix pytext tutorial syntax (#971)

Use torch.equal() instead of "==" in Custom Tensorizer tutorial (#939)

Remove and mock doc dependencies because readthedocs is OOM (#983)

Fix Circle CI build_docs error (#959)

Add OSS integration tests: DocNN (#1021)

Print model into the output log (#1127)

Migrate pytext/utils/torch.py logic into pytext/torchscript/ for long term maintainability (#1082)

Demo datasource fix + cleanup (#994)

Documentation on the config files and config-related commands (#984)

Config adapter old data handler helper (#943)

Nicer gen_config_impl (#944)

Deprecated Features

Remove DocModel_Deprecated (#916)

Remove RNNGParser_Deprecated, SemanticParsingTask_Deprecated, SemanticParsingCppTask_Deprecate, RnngJitTask,

Remove QueryDocumentTask_Deprecated(#926)

Remove LMTask_Deprecated and LMLSTM_Deprecated (#882)

CompositionDataHandler to fb/deprecated (#963)

Delete deprecated Word Tagging tasks, models and data handlers (#910)

Bug Fixes

Fix caffe2 predict (#1103)

Fix bug when tensorizer is not defined (#1169)

Fix multitask metric reporter for lr logging (#1164)

Fix broken gradients logging and add lr logging to tensorboard (#1158)

Minor fix in blockwise sparsifier (#1130)

Fix clip_grad_norm API (#1143)

Fix for roberta squad tensorizer (#1137)

Fix multilabel metric reporter (#1115)

Fixed prepare_input in tensorizer (#1102)

Fix unk bug in exported model (#1076)

Fp16 fixes for byte-lstm and distillation (#1059)

Fix clip_grad_norm_ if grad_norm > max_norm > 0: TypeError: '>' not supported between instances of 'float' and 'NoneType' (#1054)

Fix context in multitask (#1040)

Fix regression in ensemble trainer caused by recent fp16 change (#1033)

ReadTheDocs OOM fix with CPU Torch (#1027)

Dimension mismatch after setting max sequence length (#1154)

Allow null learning rate (#1156)

Don't fail on 0 input (#1104)

Remove side effect during pickling PickleableGPT2BPEEncoder

Set onnx==1.5.0 to fix CircleCI build temporarily (#1014)

Complete training loop gracefully even if no timing is reported (#1128)

Propagate min_freq for vocab correctly (#907)

Fix gen-default-config with Model param (#917)

Fix torchscript export for PyText modules (#1125)

Fix label_weights in DocModel (#1081)

Fix label_weights in bert models (#1100)

Fix config issues with Python 3.7 (#1066)

Temporary fix for Fairseq dependency (#1026)

Fix MultipleData by making tensorizers able to initialize from multiple data sources (#972)

Fix bug in copy_unk (#964)

Division by Zero bug in MLM Metric Reporter (#968)

Source code(tar.gz)
Source code(zip)
pytext-nlp-0.3.0.tar.gz(288.70 KB)
pytext_nlp-0.3.0-py3-none-any.whl(407.89 KB)
v0.2.2(Aug 15, 2019)
Note: this is the last release with _Deprecated classes. Those classes will be removed in the next release.

New Features:

DeepCNN Representation for word tagging

Combine KLDivergenceBCELoss with SoftHardBCELoss and F.cross_entropy() in CrossEntropyLoss (#689)

add dense feature support for doc model (#710)

add torchscript quantizaiton support in pytext

pytext multi-label support (#731)

open source transformer representations (#736)

open source transformer based models - data, tensorizers and tokenizer (#708)

Create AlternatingRandomizedBatchSampler (#737)

open source MaskedLM and BERT models (#734)

Support bytes input in word tagging model OSS (#745)

open source extractive question answering models (#742)

torchscriptify for ensemle task

enabled lmlstm labels exporting (#767)

Enable dense features in ByteTokensDocumentModel (#763)

created bilstm dropout condition (#769)

enabled lmlstm caffe2 exporting (#766)

PolynomialDecayScheduler (#791)

removed bilstm dependence on seq_lengths (#776)

fp16 optimizer (#782)

Add Dense Feature Normalization to FloatListTensorizer and DocModel (#859)

Add Sparsifier component to PyText and L0-projection based sparsifier (#860)

implemented cnn pooling for doc classification (#872)

implemented bottleneck separable convolutions (#855)

Add eps to Adam (#858)

implemented mobile exporter (#785)

support starting training from saved checkpoint (#824)

implemented separable convolutions (#830)

implemented gelu activations (#829)

implemented causal convolutions (#811)

implemented dilation for convolutions (#810)

created weight norm option (#809)

Ordered Neuron LSTM (#854)

Add PersonalizedByteDocModel (#816)

CNN based language models (#827)

improve csv support in TSVDataSource (#777)

Change default batch sampler DisjointMultitaskData to RoundRobinBatchSampler (#802)

Support using serialized pretrained embedding file (#797)

Documentation / Usability / Logging:

Fewer out-of-vocab print messages, with some stats (#697)

Echo epoch number to console while training (#712)

Separate timing for prediction and metric calculation. (#738)

multi-label soft metrics (#754)

changed lm metric reporting (#765)

fix data source tutorial (#762)

fix doc sphinx deprecation warning (#775)

Add the ability to pass parameter values to gen-default-config (#856)

Remove "pytext/" from paths in demo json config (#878)

New documentation about hacking pytext and dealing with github. (#862)

install_deps supports updates (#863)

Reduce number of PEP print (#861)

better error message for config with unknown component (#801)

Add Recall at Precision Thresholds to Config (#792)

implemented perplexity reductions for lm score reporting (#799)

adapt prediction workflow to new design (#746)

Bug fixes:

block sharded tsv eval/test fix (#698)

Fix BoundaryPooling tracing (#713)

fixes LMLSTM weight tying bug (#704)

Fix duplicate entries in vocab (#721)

Bugfix for trainer not reporting eval results (#740)

Reintroduce metrics export in new task (#748)

fix open source tests (#750)

Fix missing init_tensorizers arg (#893)

Fix intent slot metric reporter not working with byte offset (#883)

Fix issue with some tensorizers still re-initializing vocab when loaded from saved state (#848)

fixed overflow error in lm reporting (#831)

fix BlockShardedTSVDataSource (#832)

v0.2.1 (skipped because of packaging issues)
Source code(tar.gz)
Source code(zip)
pytext-nlp-0.2.2.tar.gz(266.37 KB)
pytext_nlp-0.2.2-py3-none-any.whl(381.24 KB)
v0.2.0(Jun 15, 2019)
Note: This release makes the new data handler API the default and deprecates Task and Model classes using the old data handler API. We recommend that you migrate your models to the new API as soon as possible. More details here: https://www.facebook.com/groups/pytext/permalink/1038962512978256/

New Stuff

most tasks and models deprecated, replaced with better versions using the new data handler API

performance improvements in metric reporter

Add Multilingual TSV Data Source

LabelSmoothedCrossEntropyLoss

Support for pretrained word embedding in TokenTensorizer

option to use pretrained embedding

TorchScript export for document classification

Improve log in trainer

performance measurement: reporting tokens_per_second and updates_per_second

Implement DocumentReader from DrQA in PyText (StackedBidirectionalRNN)

improved and updated documentation

Implement SWA(SGD|ADAM) and Adagrad Optimizers

cache numerized data in memory

TorchScript BPE tokenization

CLI command to update configs

Visualize gradients with tensorboard

Many bug fixes and code clean-ups
Source code(tar.gz)
Source code(zip)
pytext-nlp-0.2.0.tar.gz(211.96 KB)
pytext_nlp-0.2.0-py3-none-any.whl(308.73 KB)
v0.1.5(Apr 16, 2019)
v0.1.5

Note: this is a last release in 0.1.x. The next release will deprecate Task and Model base classes and make the improved API of the new data handler the default. You can start using it already by inheriting from NewTask. NewDocumentClassification and NewWordTaggingTask use this new API, and you can get the first example in the tutorial "Custom Data Format".

New Stuff

add config adapter

PyText is very young and its API is still in flux, making the config files brittle

config files now have a version number reflecting the API at the time it was created

older versions can be loaded and internally transformed into newer versions

better metrics and reporting

better training time tracking

cool new visualization of model state in TensorBoard

pretty results in the terminal

improved distributed training

torchscript export

support for SQuAD dataset

add AugmentedLSTM

add dense features support

new plugin system: command line option --include to import custom user classes (see tutorial "Custom Data Format" for example)

Many bug fixes and code clean-ups
Source code(tar.gz)
Source code(zip)
pytext-nlp-0.1.5.tar.gz(176.07 KB)
pytext_nlp-0.1.5-py3-none-any.whl(268.69 KB)
v0.1.4(Feb 7, 2019)
v0.1.4

New Stuff

Used pip freeze to pin all dependency versions as part of release, should increase stability

Refactor Metric Reporters to reduce coupling

RNNG Improvements:

Support Pretrained embeddings in RNNG

Support GPU Training

More Test Coverage

Tensorboard Support

Added QueryDocumentPairwiseRankingModel

Distributed Training Improvments:

Sharded Data Loading to reduce memory consumption

Fix Several issues with race conditions and unserializable state

Reduced GPU memory Consumption by skipping gradient computation on evaluation

And lots of bug fixes

Known Issues

PyText doesn't work with the new ONNX v1.4.0, so we have pinned it to 1.3.0 for now
Source code(tar.gz)
Source code(zip)
pytext-nlp-0.1.4.tar.gz(130.13 KB)
pytext_nlp-0.1.4-py3-none-any.whl(205.29 KB)
v0.1.0(Dec 14, 2018)
Production ready models for various NLP/NLU tasks:

Text classifiers

Yoon Kim (2014): Convolutional Neural Networks for Sentence Classification

Lin et al. (2017): A Structured Self-attentive Sentence Embedding

Sequence taggers

Lample et al. (2016): Neural Architectures for Named Entity Recognition

Joint intent-slot model

Zhang et al. (2016): A Joint Model of Intent Determination and Slot Filling for Spoken Language Understanding

Contextual intent-slot models

Distributed-training support built on the new C10d backend in PyTorch 1.0

Extensible components that allows easy creation of new models and tasks

Reference implementation and a pretrained model for the paper: Gupta et al. (2018): Semantic Parsing for Task Oriented Dialog using Hierarchical Representations

Ensemble training support

Source code(tar.gz)
Source code(zip)