decaNLP - The Natural Language Decathlon: A Multitask Challenge for NLP

•    Python

---

The Natural Language Decathlon is a multitask challenge that spans ten tasks: question answering (SQuAD), machine translation (IWSLT), summarization (CNN/DM), natural language inference (MNLI), sentiment analysis (SST), semantic role labeling(QA‑SRL), zero-shot relation extraction (QA‑ZRE), goal-oriented dialogue (WOZ, semantic parsing (WikiSQL), and commonsense reasoning (MWSC). Each task is cast as question answering, which makes it possible to use our new Multitask Question Answering Network (MQAN). This model jointly learns all tasks in decaNLP without any task-specific modules or parameters in the multitask setting. For a more thorough introduction to decaNLP and the tasks, see the main website, our blog post, or the paper. While the research direction associated with this repository focused on multitask learning, the framework itself is designed in a way that should make single-task training, transfer learning, and zero-shot evaluation simple. Similarly, the paper focused on multitask learning as a form of question answering, but this framework can be easily adapted for different approached to single-task or multitask learning.

deep-learning natural-language-processing multitask-learning

question_generation - Neural question generation using transformers

•    Jupyter

---

Question generation is the task of automatically generating questions from a text paragraph. The most straight-forward way for this is answer aware question generation. In answer aware question generation the model is presented with the answer and the passage and asked to generate a question for that answer by considering the passage context. While there are many papers available for QG task, it's still not as mainstream as QA. One of the reasons is most of the earlier papers use complicated models/processing pipelines and have no pre-trained models available. Few recent papers, specifically UniLM and ProphetNet have SOTA pre-trained weights availble for QG but the usage seems quite complicated. This project is aimed as an open source study on question generation with pre-trained transformers (specifically seq-2-seq models) using straight-forward end-to-end methods without much complicated pipelines. The goal is to provide simplified data processing and training scripts and easy to use pipelines for inference.

nlp natural-language-processing deep-learning transformer natural-language-generation nlg question-generation t5

Apache Hive - The Apache Hive (TM) data warehouse software facilitates querying and managing large d

•    Java

---

The Apache Hive (TM) data warehouse software facilitates querying and managing large datasets residing in distributed storage.

Ephyra - Question Answering System

•    Java

---

Ephyra is a modular and extensible framework for open domain question answering (QA). The system retrieves accurate answers to natural language questions from the Web and other sources. The goal is to give researchers the opportunity to develop new QA techniques without worrying about the end-to-end system.

question answer information-analysis data-analysis analysis knowledge semantic-search

tapas - End-to-end neural table-text understanding models.

•    Python

---

Code and checkpoints for training the transformer-based Table QA models introduced in the paper TAPAS: Weakly Supervised Table Parsing via Pre-training. The easiest way to try out TAPAS with free GPU/TPU is in our Colab, which shows how to do predictions on SQA.

tensorflow question-answering nlp-machine-learning table-parsing

spago - Self-contained Machine Learning and Natural Language Processing library in Go

•    Go

---

A Machine Learning library written in pure Go designed to support relevant neural architectures in Natural Language Processing. spaGO is self-contained, in that it uses its own lightweight computational graph framework for both training and inference, easy to understand from start to finish.

nlp machine-learning natural-language-processing deep-learning neural-network automatic-differentiation artificial-intelligence recurrent-networks lstm computation-graph question-answering bart automatic-translation deeplearning language-model bert transformer-architecture bert-as-service named-entities-recognition

Sequence-Semantic-Embedding - Tools and recipes to train deep learning models and build services for NLP tasks such as text classification, semantic search ranking and recall fetching, cross-lingual information retrieval, and question answering etc

•    Python

---

SSE(Sequence Semantic Embedding) is an encoder framework toolkit for natural language processing related tasks. It's implemented in TensorFlow by leveraging TF's convenient deep learning blocks like DNN/CNN/LSTM etc. Depending on each specific task, similar semantic meanings can have different definitions. For example, in the category classification task, similar semantic meanings means that for each correct pair of (listing-title, category), the SSE of listing-title is close to the SSE of corresponding category. While in the information retrieval task, similar semantic meaning means for each relevant pair of (query, document), the SSE of query is close to the SSE of relevant document. While in the question answering task, the SSE of question is close to the SSE of correct answers.

PHPMyAdmin

•    PHP

---

phpMyAdmin is a free software tool written in PHP intended to handle the administration of MySQL over the World Wide Web. phpMyAdmin supports a wide range of operations with MySQL. The most frequently used operations are supported by the user interface (managing databases, tables, fields, relations, indexes, users, permissions, etc), while you still have the ability to directly execute any SQL statement.

database-front-end adminstration tool mysql-admin mysql-management mysql-tool

transformers - 🤗Transformers: State-of-the-art Natural Language Processing for Pytorch, TensorFlow, and JAX

•    Python

---

🤗 Transformers provides thousands of pretrained models to perform tasks on texts such as classification, information extraction, question answering, summarization, translation, text generation and more in over 100 languages. Its aim is to make cutting-edge NLP easier to use for everyone. 🤗 Transformers provides APIs to quickly download and use those pretrained models on a given text, fine-tune them on your own datasets and then share them with the community on our model hub. At the same time, each python module defining an architecture is fully standalone and can be modified to enable quick research experiments.

nlp natural-language-processing tensorflow pytorch transformer speech-recognition seq2seq flax gpt pretrained-models language-models natural-language-generation nlp-library language-model bert natural-language-understanding jax xlnet pytorch-transformers model-hub

Sequel Pro - MySQL Database management for Mac

•    PHP

---

Sequel Pro is a fast, easy-to-use Mac database management application for working with MySQL databases. It gives you direct access to your MySQL databases on local and remote server. It supports Full table management including indexes and MySQL Views.

database-front-end adminstration tool mysql-front-end mysql-gui database-management

Coolstorage - ORM library for .NET

•    CSharp

---

The main strength of Vici CoolStorage is the ease of use. Most ORM tools still require a lot of unneeded code to accomplish basic data persistence tasks, but Vici CoolStorage is designed to relieve the programmer from these tedious and error-prone tasks, making it very intuitive to use.

orm persistence database-development datamapper ormdotnet

csv-schema - Analyzes a CSV file and generates database table schema, all within the browser

•    Javascript

---

This application parses CSV files (including huge ones) within the browser. It analyzes each field to suggest the best database field type, max length, and whether or not there are any null values. From there, you can rename fields, ignore them, override field types/lengths, etc. and generate database table creation sql for MySQL, MariaDB, Postres, Oracle, or SQLite3. This application uses PapaParse for CSV parsing and Knex.js for SQL query building.

NHibernate - object-relational mapper for .NET

•    CSharp

---

NHibernate is a mature, open source object-relational mapper for the .NET framework. NHibernate is a port of Hibernate Core for Java to the .NET Framework. It handles persisting plain .NET objects to and from an underlying relational database.

orm persistence database-development datamapper ormdotnet

join-monster - A GraphQL to SQL query execution layer for query planning and batch data fetching.

•    Javascript

---

Efficient query planning and data fetching for SQL. Use JOINs and/or batched requests to retrieve all your data. It takes a GraphQL query and your schema and automatically generates the SQL. Send that SQL to your database and get back all the data needed to resolve with only one or a few round-trips to the database. ...and get back correctly hydrated data.

graphql graphql-js sql nodejs sql-query join query batch data schema

CockroachDB - Cloud-native SQL database.

•    Go

---

CockroachDB is a cloud-native SQL database for building global, scalable cloud services that survive disasters.CockroachDB is a distributed SQL database built on a transactional and strongly-consistent key-value store. It scales horizontally; survives disk, machine, rack, and even datacenter failures with minimal latency disruption and no manual intervention; supports strongly-consistent ACID transactions; and provides a familiar SQL API for structuring, manipulating, and querying data.

database sql distributed-database cockroachdb newsql oltp google-spanner key-value

ClickHouse - Column-oriented database management system that allows generating analytical data reports in real time

•    C++

---

ClickHouse is a fast open-source OLAP database management system. It is column-oriented and allows to generate analytical reports using SQL queries in real-time. Large queries are parallelized using multiple cores, taking all the necessary resources available on the current server. It supports distributed processing, data can reside on different shards. Each shard can be a group of replicas used for fault tolerance. All shards are used to run a query in parallel, transparently for the user.

database columnar-database column-oriented analytics dbms olap distributed-database mpp hacktoberfest sql big-data

spaCy - 💫 Industrial-strength Natural Language Processing (NLP) with Python and Cython

•    Python

---

spaCy is a library for advanced Natural Language Processing in Python and Cython. It's built on the very latest research, and was designed from day one to be used in real products. spaCy comes with pre-trained statistical models and word vectors, and currently supports tokenization for 20+ languages. It features the fastest syntactic parser in the world, convolutional neural network models for tagging, parsing and named entity recognition and easy deep learning integration. It's commercial open-source software, released under the MIT license. 💫 Version 2.0 out now! Check out the new features here.

natural-language-processing data-science big-data machine-learning cython nlp artificial-intelligence ai spacy nlp-library neural-network neural-networks deep-learning

upscaledb - A very fast lightweight embedded database engine with a built-in query language.

•    C++

---

upscaledb is a fast key-value database which optimizes storage and algorithms for your specific data types. upscaledb can be used to build all functions of a typical SQL database, tailored to the specific needs of your application, and directly linked into your program.

nosql database key-value-store key-value embedded-database

rdb - Javascript ORM

•    Javascript

---

ORM for nodejs. Supports postgres, mySql and sqlite. 1.7.1 Support for schemas (postgres only). 1.7.0 sqlite3 is now a peer dependency. Add it to your own package.json if you intend to use it. 1.6.9 Bugfix: one-to-many relation returns empty if strategy is included. 1.6.8 Bugfix: one-to-many relation returns empty if insert/update is done earlier in transaction. 1.6.7 Bugfix in relations. 1.6.6 Bugfix. 1.6.5 Improved performance on relations. 1.6.4 Bugfix. 1.6.3 Bugfix: potential incorrect timeZoneOffset when serializing date to JSON. Got timeZoneOffset from now() instead of on actual date. 1.6.2 Removed es6 syntax to ensure backwards compatability. Fixed global var leak. 1.6.1 Now supporting sqlite. 1.6.0 Bugfix: potential ambigous column error when using limit and relating to other tables. 1.5.9 Bugfix: using multipleStatements in mySql could sometimes cause an error when updates are run right before a select. Improved performance on limit when relating to other tables. Using uuid instead of node-uuid Updated all dependencies but generic-pool to latest. (Generic-pool has some breaking changes in latest. I will update it in next release.) 1.5.8 Cleanup line breaks in documentation. 1.5.7 Bugfix: getById.exclusive and tryGetById.exclusive did not lock if row was cached. Improved performance on tryGetFirst. 1.5.6 Raw sql filters can accept sql both as string and as function. E.g. var filter = {sql: function() {return 'foo > 1';}}. 1.5.5 Optional locks for getMany, tryGetFirst and tryGetById. Instead of calling getMany(params) just call getMany.exclusive(params). Same syntax goes for tryGetFirst and tryGetById. This will result in SELECT FOR UPDATE. Bugfix: bulk deletes now accepts raw sql filters too. 1.5.4 Transaction locks. Postgres only. 1.5.3 Upgraded to pg 6.0.3 1.5.2 Improved performance and reduced memory footprint. 1.5.1 Documented JSON column type. Bug fix: Insert and foreign key violation. 1.5.0 JSON column type. Postgres json type does not support rdb filters. 1.4.1 Empty filter would sometimes cause invalid filter. 1.4.0 Raw SQL query. 1.3.0 getMany() now supports limit and orderBy - same syntax as in streaming. 1.2.3 Bugfix: iEqual gave incorrect sql when parameterized. 1.2.2 Exlusive no longer returns a clone of table. It has changes current table to exclusive locking. 1.2.1 Bugfix: Exclusive row locks 1.2.0 Exclusive row locks 1.1.0 Now supporting streaming. Requires postgres or MySQL >=5.7.7 1.0.8 README fixup. 1.0.7 Better performance on insert and update. 1.0.6 Bugfix: Transaction domain should not forward rdb singleton from old domain. 1.0.5 Documentation cleanup. 1.0.4 orderBy in toDto(). 1.0.3 toDto() using next tick on every thousandth row to avoid maximum call stack size exceeded. 1.0.2 Reduced number of simultaneous promises in order to avoid maximum call stack size exceeded. 1.0.1 Bugfix: Incorrect insert/updates on timestamp without timezone. The time was converted utc instead of stripping the timezone. 1.0.0 Transaction domain forwards properties from old domain. Semantic versioning from now on. 0.5.1 Improved performance 0.5.0 Logging: rdb.log(someFunc) logs sql and parameters. Raw sql filters. 0.4.9 New method: tryGetById. New filter: iEqual, postgres only. Bugfix: rows.toJSON() without strategy did not include any children. 0.4.8 Explicit pooling with size and end(). Bugfix: mySql did not release client to pool. 0.4.7 Upgraded to pg 4.3.0 Upgraded to mysql 2.5.5 0.4.6 Upgraded pg 4.2.0. 0.4.5 Oops. Forgot to use pg.js instead of pg. 0.4.4 Upgraded all dependencies to latest. Using pg.js instead of pg. 0.4.3 Can ignore columns when serializing to dto. 0.4.2 Bugfix: update on a row crashes when a delete occurs earlier in same transaction. 0.4.1 Bugfix: more global leaks. 0.4.0 Bugfix: global leak. 0.3.9 Bugfix: eager loading joins/hasOne with non unique column names was not handled correctly. 0.3.8 Supports mySql. Bulk deletes. 0.3.7 Bugfix: eager loading manyRelation on a join/hasOne returned empty array #11 0.3.6 Fixed sql injection vulnerability. 0.3.5 Built-in fetching strategies for lazy loading. Works best in readonly scenarios. 0.3.4 Docs and examples split moved to separate file. 0.3.3 Fixed documentation layout again. 0.3.2 Fixed documentation layout. 0.3.1 Case insensitive filters: iStartsWith, iEndsWith and iContains. 0.3.0 Fix broken links in docs. 0.2.9 Support for row.delete(). Rollback only throws when error is present. 0.2.8 Guid accepts uppercase letters. Bugfix: null inserts on guid columns yielded wrong sql. 0.2.7 New method, toDto(), converts row to data transfer object. Bugfix: toJSON returned incorrect string on hasMany relations. 0.2.6 Fixed incorrect links in README. 0.2.5 Bugfix: caching on composite keys could give a crash #7. Improved sql compression on insert/update. 0.2.4 Bugfix: getMany with many-strategy and shallowFilter yields incorrect query #6. 0.2.3 Reformatted documentation. No code changes.

orm sql mysql postgres pg

lectures - Oxford Deep NLP 2017 course

•    

---

This repository contains the lecture slides and course description for the Deep Natural Language Processing course offered in Hilary Term 2017 at the University of Oxford. This is an applied course focussing on recent advances in analysing and generating speech and text using recurrent neural networks. We introduce the mathematical definitions of the relevant machine learning models and derive their associated optimisation algorithms. The course covers a range of applications of neural networks in NLP including analysing latent dimensions in text, transcribing speech to text, translating between languages, and answering questions. These topics are organised into three high level themes forming a progression from understanding the use of neural networks for sequential language modelling, to understanding their use as conditional language models for transduction tasks, and finally to approaches employing these techniques in combination with other mechanisms for advanced applications. Throughout the course the practical implementation of such models on CPU and GPU hardware is also discussed.

deep-learning machine-learning natural-language-processing nlp oxford