Find open source projects actively accepting contributors. Search repositories, filter by program milestones, difficulty tags, or tech stack.
Use our Orbit AI Matcher to find out! Get instant matching scores based on your developer skills, preferred frameworks, and contribution experience.
Convert your selected open-source project into a winning GSoC, LFX, or Outreachy application using Proposal Studio.
In March 2016, AlphaGo became the first computer Go program to beat one of the best go players in the world. But Deepmind didn’t stop there and kept developing better versions of it, namely: AlphaGo Zero, AlphaZero, and MuZero. AlphaZero.jl is an open-source implementation of one of those and became a great showcase of Julia’s unique strengths. And it is the 8th most starred Julia package. This Google Summer of Codes aims to add on top of AlphaZero, the MuZero implementation into AlphaZero.jl and create an interactive and pedagogical website that not only explains the MuZero algorithm but also illustrates deeper explanations of its hyperparameter tuning and implementation tradeoffs. This GSoC will not only expand a great Julia package but also give more visibility and illustrate Julia’s advantages in Machine Learning projects.
I will potentially work towards the following problems during the GSoC: 1) I am excited to implement advanced algorithms in JAX and create/re-produce demos/figures to put in textbooks. For example, recently I implemented Automatic Differentiation Variational Inference (ADVI) from scratch in JAX. Readers can easily connect maths with code by referring to this kind of from-scratch implementation. Also, pure JAX code can introduce some unique features of JAX such as lax.scan() to optimize the training loop. 2) I will improve the quality of the existing figures in textbooks using a set of matplotlib configurations we use in our group for research papers. For example, In Figure 2.4 of book1, due to missing y-label, x-label, and legend, readers may find difficulty in understanding the figure. I will fill these small gaps and also improve the figure by applying latexify on them. Latexify is a function that applies latex effects by setting consistent font style, font size, and figure size with other text in the book. For example, this gist (https://gist.github.com/karm-patel/9c5c2ee77afd495759ed1a2aad0e460e) shows an improvement of the above figure. I have already started to work on this thread and contributed by improving Figure 2.2, Figure 2.4, Figure 2.5, Figure 2.6, Figure 2.9, Figure 2.11, and Figure 2.15 of the book1. 3) Add an Interactive demo for figures wherever applicable: I will add some interactive demos for figures so the reader can visualize the effect of variables in a particular problem. For example, figure 2.9 shows two different binomial distributions for two different θ (probability of success). I have added an interactive demo as shown in this notebook (https://colab.research.google.com/github/probml/pyprobml/blob/master/notebooks/book1/02/binom_dist_plot.ipynb) In above notebook readers can change θ and understand the effect of θ on the binomial distribution.
F2PY is an open-source utility that provides an easy connection between Python and Fortran languages. It enables users to integrate Fortran code into Python easily and quickly. F2PY has been in maintenance mode since 2009, ever since it became feature-complete for Fortran 77. Beginners browsing its codebase in the current state might find it challenging to read, understand and contribute. The frontend of F2PY is a handwritten command line parser that predates the intrinsic ‘argparse’ module by several years. This project will include modernising the CLI with the “argparse” python library, simplifying the frontend codebase and making it more developer-friendly to contribute to. Since “np.distutils” is set for deprecation, I propose adding an option to build the F2PY’s generated CPython extension modules with Meson. Meson will significantly speed up building standalone extension modules through the F2PY CLI. There is no change proposed to the process of generating C and Fortran wrappers from Fortran source files. Generation of the source code for integration in Python projects will be carried out in the same way. Currently, “np.distutils” is used by F2PY to build generated extension modules. Adding a Meson backend will provide faster builds and long-term backend support for F2PY. Additionally, I plan to re-implement the class-based test-suite of F2PY in a modern pytest manner and add a developer’s guide to F2PY.
Ganga routinely handles a tremendous amount of computational tasks simultaneously and keeping track of their status is prone to lagging behind with its current implementation, particularly regarding remote jobs submitted through DIRAC. Furthermore, it is the project mentors’ belief that the current implementation of the monitoring service could be simplified and brought up to date. My goal at the end of the project’s timeline would be to deliver a new and efficient monitoring service using the asyncio framework. The main goal is for each backend’s monitoring to be rewritten using asyncio to make it as efficient as possible, but the service will also support multithreading and multiprocessing executors. It will be backend agnostic and allow any new backend to easily hook up to it. I also plan to specifically address DIRAC’s current inefficient monitoring implementation by using its REST API to retrieve running job information.
DBpedia provides monthly releases produced by the DBpedia Extraction Framework. They are composed of various data artifacts that mainly stem from the wiki dumps. However, some of them also rely on API calls for rendering dynamic contents, which is the case of the DBpedia abstracts. The large amount of data requested from APIs couldn't be extracted entirely within a month today. We suggest solving this issue by a strategy composed of four steps: - a study based on the data recorded during the last abstract extraction - the test and implement the use of the TextExtracts extension and the improvement the error management - the reduction the number of possible calls - the integration into the framework of the possibility to appeal to more than one API Each step of the project will be developed into a new dedicated GitHub branch of the DBpedia extractor framework, which could be documented and used for working on the project.
People love to listen to music together, for which they would like to start a video conference where they can enjoy music together with their friends and family. A music bot for Jitsi video meetings will serve this purpose. This Music bot will be extremely simple to use, while in a meeting, just copy the meet link to the bot window and start the bot. The bot joins the meeting as a participant and follows the commands. The bot will be controlled by writing simple English commands in the chat box like /play <songName>, /pause, /resume, etc. User can also use their accounts on music streaming apps with the music bot to provide them with high-quality ad-free music if they enjoy a premium subscription on their accounts.
ColorPicker has RGB, HSV and Raw mode, but they never were properly separated, resulting in messy code. With OKHSL mode going to be added soon that adds even more messy code. The aim of this project is to make code in ColorPicker refactored and cleaned up to make multiple color modes more maintainable and easier to add, also to update the UX of ColorPicker.
Refactor workbench to a React + Typescript implementation and improve various sections of the application including Table view, file uploads, data sync across sections, etc
The goal is to support a fully emulated UEFI environnement in QEMU using the MinPlatform specification. Deliverables would correspond to the different stages of the Minimum Platform implementation. The first one would be a silicon specific package for QEMU-x86-64 The first objective would be to get an UEFI shell inside of qemu-x86-64, and then getting an OS to boot. If time allows it, getting OpenQEMUBoardPkg to run on additional architecture could be another objective. The first month could be used to get to a working UEFI shell, the second to add the missing features to get a feature-complete package for UEFI emulation, and if everything goes as planned, porting to other Qemu-supported architectures. FSP Stretch goal can be achieved by porting the existing QemuFSP package, reducing complexity for platform init on Intel platforms (which qemu-system-x86_64 emulates).
The goal of this project is to migrate the exploration editor page from angularJs to angular. Users will be able to experience faster page transitions and a better overall user experience after this project is completed. This project also involves switching from webpack to angular CLI, Because of its simplicity and features, migration to angular CLI will improve the entire development experience. After completion of the project, new programmers starting with oppia will be able to grasp the codebase easily.
The ever-growing body of data in Neuroscience, which has been enabled by the advancement of state-of-the-art techniques, calls forth collaborative attempts to make data FAIR (findable, accessible, interoperable and reusable). Consequentially, Neurodata Without Borders (NWB), an open-source software that defines components of a language for neurophysiology data, was born out of such need. NWB has thus exemplified as a sustainable biological data ecosystem, but there lies the need of converting readily available datasets into NWB format to allow for further sharibility and producibility. With that said, this project aims to (i) convert a publically available dataset into NWB format, (ii) ensure NWB datasets' compatibility with NWB explorer (NWBE), (iii) and facilitate generality of the conversion process by investigating the sources of NWBE-NWB incompatibilities and ensuring essential metadata being added. This is a medium size project (175h).
# proposal summary PostgreSQL is a very actively developed project and there are so many existing codes and so many new ones are added. I think there are still many functions that have not been tested. Test codes that various cases increase development stability and other developers can make changes safely. I will write a test code for the module that has not been tested yet. # list of modules to test - pg_dump - psql - pg_ctl, - initdb - pg_stat_statments - pg_buffercache - pg_prewarm - pgrowlocks - pg_stattuple - pg_waldump - pg_basebackup
TensorFlow, Keras, and PyTorch are the most prominent libraries used for building DL and ML models. The use of PyTorch, as the primary backend for building algorithms, is growing due to its sustainability. Lately, DeepChem has decided to port some TensorFlow models to PyTorch. This project aims to successfully migrate one of the organization's 20 models. This model will perform a mix of invertible transformations between the base and target distribution. In order to optimize the results, a Normalizing Flow object initialized with the flow layers and base distribution would be iterated computing the loss for each epoch. For this reason, transformations and flows layers will be created with their respective unittests. Finally, the implementation will include the respective documentation and a tutorial, if applicable, following DeepChem's API.
Sugar Labs has more than 250 activities GitHub and elsewhere which have scope for improvement. Since the support for Python2 was withdrawn from Python Foundation, porting the activities to Python3 and GTK+ 3 is very crucial. My work on this project includes, (1) Porting activities to GTK+ 3 and Python3. (2) Implementing basic game design features which attracts the elementary grade children (activities from Honey). (3) Adding activity specific features like collaborations and enhance User Interface & User Experience. (4) Testing and modifying the activities to ensure 12 activities are ready to release.
With more than 1 million new diagnoses reported every year, prostate cancer (PCa) is the second most common cancer among males worldwide that results in more than 350,000 deaths annually. The key to decreasing mortality is developing more precise diagnostics. Keeping this in mind, this project attempts to develop a sophisticated diagnosing technique that will highlight the potential cancerous regions in a given ‘Prostate biopsy slide’, thereby aiding the doctors in faster and more accurate diagnosis. However, in the presence of a Subject Matter Expert (SME) (doctors, radiologists, etc.…) this task becomes trivial, but in the absence of SMEs diagnosing PCa via 'Whole Slide Images' is almost always impossible. In many developing countries, the healthcare sector is drastically underfunded and the vast majority of these countries cannot provide basic healthcare to the public. In such conditions, the availability of doctors is very unlikely. Automated analysis and classification of biopsy images without the presence of any subject matter expert is a crucial task. Thus, developing a reliable and robust system, susceptible to noise, in the form of a web app or inference engine is the project's main goal.
A knowledge extraction pipeline for artworks from christian iconography that will be used to create a knowledge graph for Christian Iconography. The intention of knowledge graph is that it should help untutored eyes find connections between different artworks. This project is divided into 4 main parts: 1) A model to find the parts of the image that might be useful to us and extract those regions of interest 2) A classification model that would extract possible classes as well as items and use the confidence scores 3) Defining an ontology that would be the skeleton for our knowledge graph. It would be able to tackle the problem of finding connections between separate artworks using the confidence scores found in the previous step 4) Finally, we would populate our knowledge base with the help of our ontology.
Render blocking nature of resources is important to pre-emptively pinpoint exactly at which point content will be rendered on screen. The render blocking status is currently determined by making use of heuristics to understand which resources block render. The project aims to document in the formal specification the conditions which a resource would block render and provide an API to know the exact classification(with regards to render-blocking) that was done.
This project aims to improve the existing deep learning for computer vision ecosystem in Julia, especially by adding more models that users can use without much modification. Metalhead.jl is the package where this project will be working on consolidating all such computer vision models., The goal will be to add models that are both customisable and extensible so that they can be used for tasks such as image classification and semantic segmentation with ease.
The current onboarding workflow that the Oppia Android app offers is not intuitive enough. Users get confused or spend more time than necessary figuring out the features and how to navigate in the app due to lack of a better onboarding workflow. Sometimes users do not use a feature available to them that can ease their learning journeys due to lack of information. To ensure first time users have the best experience possible and to create an onboarding experience that is: Engaging but not time-consuming Educational but not text-heavy Interactive but not complicated Does not take too much time or clicks, especially if it is not interactive Does not add unnecessary visual or cognitive complexity Easily dismissible at all parts of the experience The app should show the onboarding flow the first time the user experiences whichever screen contains the interactive elements that should be explained. Each of these experiences may lead to multiple parts of the UI being highlighted, and should follow a predefined linear flow. There should be an option to cancel out of the flow, and the app should not show it again if the flow is either canceled or completed. Users should be able to access the onboarding flow if they cancel accidentally or forget something and want to go over it again.
Kolibri lacks representable authentic testing data , also generating testing data utilities are run during unit test runtime making them not so efficient. This project aims at 1- Improving Kolibri Testing Data for its various apps by generating ready to load staic fixtures files which should be authentic,more relevant to real usage and comprehensible. 2- Wiriting Scripts responsible for generating determinstic data on the fly so developers writing test cases get to choose suitable options based on the current testing scenario. 3- Writing Scripts to update Fixtures files according to changes in the databse models.
ENIGMA is a free, open source development environment and game engine. It supports game development in Linux, Windows and MacOS and also Android as an endpoint. In order to support these numerous platforms a good testing suite is required so that ENIGMA developers can be productive instead of worrying about their code breaking on a platform they are not familiar with. As of yet the test harness runs its numerous tests only on linux. This proposal aims to provide a win32 based port of the x11 test harness and integrate it with azure pipelines which will allow the engine to be thoroughly tested in both linux and windows.
In this project, we aim to implement piecewise function on the polyhedral complex for general purpose for SageMath. Generally, this project should achieve efficient symbolic and numerical evaluation of piecewise functions given sub-domains and functional coefficients provided by users. Considering the computational complexity, efficient piece searching algorithm should also be designed, especially for the case of a large number of pieces being provided. Besides general purposes, improvements would also be considered with specific purposes, if provided, of piecewise linear function including but not limited to approximation, evaluation of definite integrals, and regression. Lastly, the whole project should be compatible and communicate well with other objects in SageMath.
On Perfetto, users can get brilliant insights into the performance of applications through trace visualizations and metrics. Recently, core web vitals shaped a different view of a smooth website by bringing the Interaction to Next Paint metric to the light. Therefore, there is a need to update the Perfetto with the latest revolutionary metrics to provide better and faster insights to the developers of their web applications. The goal of this project is to improve the Perfetto tracing tool and visualize the core web vital metrics and INP in a consistent manner. Make Perfetto Perfect!
TiSpark maintains a com.pingcap.tikv/tikv-client module. It is a fork of the current official tikv-client legacy version. The difference from the current official tikv-client is that some TiSpark-specific enhancements have been added. This makes enhancements, bug fixes made for TiSpark tikv-client not affecting the official tikv-client and vice versa. We want to normalize the Java client and use the official one. Do not break original functionality and guarantee consistent behavior.