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.
This project addresses the lack of accessible, Greek high-quality datasets for training Large Language Models (LLMs). Although robust LLM frameworks are widely available, open-source Greek-specific datasets are severely underrepresented, especially those that captures regional variations and dialects. This project's mission is to integrate Optical Character Recognition (OCR) capabilities into the processing pipeline and expand its dataset resources to support these linguistic nuances.
GIMPs typographic engine is currently limited by legacy GTK3 and Pango constraints. This project overhauls the text backend by directly integrating HarfBuzz to extract font data and advancing GIMPs experimental text layout library. Deliverables include migrating the font chooser to modern GTK list models, grouping fonts by WWS families, and introducing modular on canvas UI controls for variable font axes and OpenType features.
<p>Guarani is one of the most widely spread indigenous languages of southern South America. It is spoken by 6 million people in Paraguay (where it is one of the official languages), Brazil, Argentina and Bolivia. Guarani translators are present online, but there is no rule-based translator with morphological analysis, which could be more plausible than translators made from Spanish/Portuguese - Guarani parallel corpora. So I believe we can improve the quality of translation by applying rule-based model.</p>
This proposal outlines the development of an AI Agent for Apache HertzBeat using the Model Context Protocol (MCP), enabling natural language interaction with monitoring capabilities. The project will implement an MCP-compliant server and client architecture integrated with SpringAI, along with a chat-based frontend for the HertzBeat dashboard. This enhancement will make HertzBeat's monitoring features more accessible while serving as a reference implementation for AI integration in monitoring tools.
`cabal check` is a tool to check the correctness of a `.cabal` file and more generally to provide more useful output (warnings, especially related to uploading the package to Hackage). As now `cabal check` is an imprecise tool, outputting spurious warnings on well formed `.cabal` files. The mistakes become more and more annoying the more complex your package is (compilation flags, etc.). Picking up an AST, transforming it appropriately and *traversing* it is the only way to practically solve these problems. This proposal aim to do so.
Move2Kube is a command-line tool for automating creation of Infrastructure as code (IaC) artifacts. It has built-in support for creating IaC artifacts for replatforming to Kubernetes/OpenShift. Currently we have support for resources such as ArgoCD, Tekton, etc. There is still a gap to be covered in the support Move2Kube provides. Example - enhance support for external transformers (GRPC, file/folder permissions, etc.). Expected Outcome: - More comprehensive support for Move2Kube advanced resources and other components.
To improve pathology data collection, caMicroscope has added specific tools for assisting with annotations and classifications. However, these tools require a pretrained model and the training workflow is difficult to use. The machine learning training assistant project aims to consolidate and improve the training workflow while also enhancing the user experience. This project may involve adding a microservice to enable training on a server and runtime analysis to determine the most useful information at a given time.
OrcaAL is an active learning tool built by past GSoC students for identifying orca calls. It achieves this by leveraging both expert and machine annotations to train better models in a semi-supervised way. Past efforts have been made mostly to develop a functional proof-of-concept and improve the performance of the tool. As such, this project aims to improve the underlying infrastructure supporting the tool, as well as enhance future developers' experience when expanding upon OrcaAL's capabilities.
<p>This proposal aims at designing, implementing and evaluating plugins that integrate RocketMQ with both the HBase data store and the Beam data processing model. These integration plugins will improve RocketMQ offline storage capabilities and benefit users with stringent large-scale and data-intensive processing needs. End results of this proposal include HBase and Beam plugin implementations, as well as exhaustive unit tests, application examples and documentation.</p>
<p>The project is to add more comprehensive support for mcmc sampler methods in PyMC4. Currently there is no support for discrete distributions in model, support for gradient-free samplers, etc. The goal of the project is to expand the support for mcmc sampling, add various samplers, including SMC, etc.. Also, from the design perspective there is a need for compound step/Gibbs sampling. We need to collaborate with the Tensorflow Probability team developers to implement the most effective solution.</p>
skpro's probabilistic regressors have limited online/streaming support: only 3 of 20+ regressors implement update(), few meta-strategies exist, and there is no model persistence. This project adds: 1) native _update() for regressors with incremental APIs (NGBoostRegressor, BayesianLinearRegressor, BaggingRegressor) 2) composable meta-strategies (OnlineSlidingWindow, OnlineBatchMixture, OnlineExponentialForgetting) usable with any regressor 3) update() propagation through Pipeline and TransformedTargetRegressor 4) save()/load() ported from sktime's serialization pattern
<p>The iregnet package is the first R package to support four types of censoring and elastic net (L1 + L2) regularization. Though it is already useful for making predictions in data sets with possibly censored observations, the current code is too slow for large data sets. After this GSOC project, its model fitting code will be even faster. The iregnet package will get better performance in large data sets.</p>
JNIgen enables the generation of Dart bindings for Java/Kotlin libraries, but writing equivalent Dart code to interact with these bindings remains challenging due to language differences. This project focuses on automating the translation of Android (Java/Kotlin) docs snippets into JNIgen-compatible Dart code using a Large Language Model (LLM). It makes Android documentation more accessible to Dart developers, allowing them to directly use and experiment with examples in their projects.
<p>The Kubernetes Multi-Tenancy Working Group is chartered with exploring and defining multi-tenancy models for Kubernetes. The Multi Tenancy Benchmarks effort focuses on measuring and validating the desired behaviors for multitenancy. Part of this effort is to automate behavioral and configuration checks on existing clusters, which will be the focus of this project. This automated test suite will help all Kubernetes users validate whether their clusters are set up correctly for multi-tenancy.</p>
<p>Map/Reduce frameworks are getting more and more popular for big data processing (for example Hadoop). By utilising the unified and standards conforming API of the HPX runtime system, we believe to be able to perfectly represent the Map/Reduce programming model. Many applications would benefit from direct support in HPX. This might include adding Hypertable or similar libraries to the mix to handle the large data sets Map/Reduce is usually used with.</p>
Data scientists need a ML platform to develop, test, and deploy ML models easily. In this project, we will design and build a self-service ML platform on top of KubeVela. We will use KubeVela to provide high level workflow and APIs to glue and simplify deployment pipelines. We will also use Cloud resources to support deployment and operations tasks like domain routing, monitoring, health checking, etc.
<p>The project involves designing an evaluation model, for assessing the relative efficiency of various reweight_by_utilisation algorithms and attempts to revamp the existing reweight_by_utilisation, so as to improve upon the general performance of Ceph. The proposal tries to address the first part, by an algorithm to develop a normalised, weighted evaluation procedure and the second part, by an algorithm that relies on an iterative feedback mechanism to improve upon the efficiency (better distribution, faster convergence and less network clog).</p>
Apache Beam is a unified programming model for authoring embarrassingly parallel batch and streaming data processing pipeline. The relatively new Beam YAML SDK was introduced in the spirit of making data processing easy, but it has gained little adoption for ML tasks and was not widely used with Kafka and Iceberg. We want to address this gap by introducing new illustrative workflows and documentation for ML use cases and integration with Kafka and Iceberg using the YAML SDK.
<p>SLOPE package offers implementations that solve the Sorted L-One Penalized Estimation (SLOPE) model for various objective functions. However when predictors are correlated, current implementation is slow because it uses FISTA. As a part of this project, we aim to create a testing suite which contains implementations different algorithms, namely ADMM, proximal-Newton methods etc. and benchmark their performance. The best one will finally be incorporate into the SLOPE package.</p>
<p>Realtime capabilities are an important part of the requirements of automotive systems. One approach is to partition your CPU using a partitioning hypervisor and run an independent RTOS with the time-sensitive code. The idea is to port the partitioning hypervisor 'jailhouse' to an AGL reference platform (Raspberry-Pi 4 model B) and demonstrate that a guest rtos can be running in a 'cell'. An extension to the guest that can communicate with AGL middleware over websocket will be added.</p>
<p>Recent research has produced impressive methods of 3D reconstruction of scenes containing non-rigid deformations using a RGB-D camera, however there is no open source code available for the described solutions. The aim of this project is to extend an available implementation of KinectFusion (Newcombe et al 2011) to implement the method described in the DynamicFusion (Newcombe et al 2015) paper, and possibly extend it to support more than a single deformable model.</p>
The objective of this project is to implement Token Classification tasks for the KerasNLP library. The focus will be on enhancing the tokenizers API to allow for greater flexibility with offset mapping use cases. The ultimate goal is to use the enhanced tokenizer for the implementation of a Token Classification Pipeline. This will involve creating separate preprocessing and task classes for all the models. Finally, the project will be supplemented with a Keras-io example demonstrating Token Classification using the KerasNLP library.
The project’s objective is to smoothly incorporate an AI module library into Joomla CMS so that site managers, developers, and content producers can take advantage of AI-powered tools for improved user engagement, content management, and creation. A few of the varied use cases that will be made possible by the integration’s adaptable framework, which supports a number of AI models and providers are: • Text Generation. • Image Creation. • Content Analysis. • Intelligent Search.
This project aims to improve the previous iteration of the project where the hypothesis of it was the existence of a relation between the body gestures and the time expressions we speak and using neural networks we try to accept or reject this hypothesis. So, I proposed a new multimodal graph-based neural network architecture to classify the body keypose points by constructing a spatio-temporal graph and audio features are fused with it to converge the model to test this hypothesis. Also, suggested transfer learning to overcome the lack of data.