Search Results for "alignment" - Page 2
Sort By:
Showing 83 open source projects for "alignment"
View related business solutions-
$300 Free Credits to Build on Google CloudPut your $300 in credit toward real workloads, then keep building with free monthly usage for 20+ products. No commitment and no charge until you upgrade.
-
Cut Data Warehouse Costs by 54%BigQuery delivers 54% lower TCO with exabyte scale and flexible pricing. Free migration tools handle the SQL translation automatically.
-
1
RLHF-Reward-Modeling
Recipes to train reward model for RLHF
...The repository provides training recipes and implementations for building reward and preference models using modern machine learning frameworks. It supports multiple optimization strategies commonly used in alignment pipelines, including reinforcement learning with PPO, iterative supervised fine-tuning using rejection sampling, and direct preference optimization methods. The project also includes evaluation results showing that the trained reward models can achieve competitive performance compared with other open-source alignment systems. -
2
InfiniteYou
Flexible Photo Recrafting While Preserving Your Identity
...Using an architecture built around diffusion transformers (DiTs), InfiniteYou introduces a component called InfuseNet that injects identity features derived from reference images into the generation process — via residual connections — so that the output matches the person’s identity closely, without sacrificing visual quality or text-image alignment. The team uses a multi-stage training strategy with synthetic multi-sample data per identity to fine-tune for both identity consistency and aesthetic quality. Compared to prior methods, InfiniteYou significantly improves on identity similarity, text-prompt adherence, overall image quality, and avoids common problems such as face copy-pasting artifacts. -
3
Hunyuan3D 2.0
High-Resolution 3D Assets Generation with Large Scale Diffusion Models
The Hunyuan3D-2 model, developed by Tencent, is designed for generating high-resolution 3D assets using large-scale diffusion models. This model offers advanced capabilities for creating detailed 3D models, including texture enhancements, multi-view shape generation, and rapid inference for real-time applications. It is particularly useful for industries requiring high-quality 3D content, such as gaming, film, and virtual reality. Hunyuan3D-2 supports various enhancements and is available... -
4
ERNIE
The official repository for ERNIE 4.5 and ERNIEKit
...The repository positions ERNIEKit as an industrial-grade development toolkit, emphasizing end-to-end workflows that span high-performance pre-training, supervised fine-tuning, and alignment. It supports both full-parameter training and parameter-efficient approaches so teams can choose between maximum quality and lower-cost adaptation depending on their constraints. The project also emphasizes optimization techniques for large-scale training, including mixed-precision and hybrid-parallel strategies that are commonly needed for multi-node GPU clusters. ... -
Demo Series - Small Business Backup By VeeamWatch this on-demand demo series and learn how to protect your Microsoft 365 data with clear, simple, actionable steps that are easy to implement for businesses of all sizes.
-
5
Heretic
Fully automatic censorship removal for language models
Heretic is an open-source Python tool that automatically removes the built-in censorship or “safety alignment” from transformer-based language models so they respond to a broader range of prompts with fewer refusals. It works by applying directional ablation techniques and a parameter optimization strategy to adjust internal model behaviors without expensive post-training or altering the core capabilities. Designed for researchers and advanced users, Heretic makes it possible to study and experiment with uncensored model responses in a reproducible, automated way. ... -
6
Chatterbox
SoTA open-source TTS
Chatterbox is Resemble AI's first production-grade open source TTS model. Licensed under MIT, Chatterbox has been benchmarked against leading closed-source systems like ElevenLabs and is consistently preferred in side-by-side evaluations. Whether you're working on memes, videos, games, or AI agents, Chatterbox brings your content to life. It's also the first open source TTS model to support emotion exaggeration control, a powerful feature that makes your voices stand out. Try it now on our... -
7
HeartMuLa
A Family of Open Sourced Music Foundation Models
HeartMuLa is the open-source library and reference implementation for the HeartMuLa family of music foundation models, designed to support both music generation and music-related understanding tasks in a cohesive stack. At the center is HeartMuLa, a music language model that generates music conditioned on inputs like lyrics and tags, with multilingual support that broadens the range of lyric-driven use cases. The project also includes HeartCodec, a music codec optimized for high... -
8
DataChain
AI-data warehouse to enrich, transform and analyze unstructured data
...Datachain can persist features of Python objects returned by AI models, and enables vectorized analytical operations over them. The typical use cases are data curation, LLM analytics and validation, image segmentation, pose detection, and GenAI alignment. Datachain is especially helpful if batch operations can be optimized – for instance, when synchronous API calls can be parallelized or where an LLM API offers batch processing. -
9
SWIFT LLM
Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs
SWIFT LLM is a comprehensive framework developed within the ModelScope ecosystem for training, fine-tuning, evaluating, and deploying large language models and multimodal models. The platform provides a full machine learning pipeline that supports tasks ranging from model pre-training to reinforcement learning alignment techniques. It integrates with popular inference engines such as vLLM and LMDeploy to accelerate deployment and runtime performance. The framework also includes support for many modern training strategies, including preference learning methods and parameter-efficient fine-tuning techniques. ms-swift is designed to work with hundreds of language and multimodal models, providing a unified environment for experimentation and production deployment. -
Build Agents and Models on One PlatformGemini Enterprise Agent Platform is Google Cloud's comprehensive platform for developers to build, scale, govern, and optimize agents and models. Choose from Google's most advanced models and third-party models like Anthropic's Claude Model Family.
-
10
DOLMA
Data and tools for generating and inspecting OLMo pre-training data
DOLMA (Data Optimization and Learning for Model Alignment) is a framework designed to manage large-scale datasets for training and fine-tuning language models efficiently. -
11
GLM-4.6
Agentic, Reasoning, and Coding (ARC) foundation models
GLM-4.6 is the latest iteration of Zhipu AI’s foundation model, delivering significant advancements over GLM-4.5. It introduces an extended 200K token context window, enabling more sophisticated long-context reasoning and agentic workflows. The model achieves superior coding performance, excelling in benchmarks and practical coding assistants such as Claude Code, Cline, Roo Code, and Kilo Code. Its reasoning capabilities have been strengthened, including improved tool usage during inference... -
12
CyberPPT
A Codex Skill for generating high-density, editable PowerPoints
...Image-generation blueprints guide reconstruction while native text, shapes, tables, charts, and vectors preserve core editability. Multiple quality gates inspect structure, visuals, overflow, spatial alignment, curves, and editable elements. Delivery includes the PPTX, rendered previews, manifests, and QA results, with failed hard gates blocking completion. -
13
Protenix
A trainable PyTorch reproduction of AlphaFold 3
Protenix is an open-source, trainable PyTorch reimplementation of AlphaFold 3, developed by ByteDance with the goal of democratizing high-accuracy protein structure prediction for computational biology and drug-discovery research. Protenix provides a complete pipeline for turning protein sequences (with optional MSA / sequence alignment) or structural inputs (e.g. PDB/CIF) into full 3D atomic-level structure predictions. It supports both “full” models and lightweight variants such as “Protenix-Mini,” offering a trade-off between speed/compute cost and predictive accuracy — making structure prediction accessible even in resource-constrained environments. The project also includes support for constraints (e.g., specifying residue- or atom-level contact constraints, or pocket constraints) to guide predictions toward biologically or experimentally relevant conformations, which enhances its utility for tasks like modeling complexes, ligands, or antibody–antigen interactions. -
14
Fish Speech
SOTA Open Source TTS
...The repository hosts the code and tooling for training, fine-tuning, and serving high-quality TTS, while the current flagship models (OpenAudio-S1 and S1-mini) are distributed via Fish Audio’s playground and Hugging Face. The models are evaluated with Seed TTS metrics and achieve exceptionally low word and character error rates, indicating strong intelligibility and alignment between text and audio. Fish Speech emphasizes expressive and controllable voices: it supports a long list of emotion tags, tone markers, and special audio effect markers that can be embedded in the text to drive prosody and vocal style, from basic emotions to nuanced states like sarcastic, conciliative, or hysterical. The system is multilingual and cross-lingual, handling multiple languages in a single input without explicit phoneme markup, and is trained on large-scale datasets. -
15
PaLM + RLHF - Pytorch
Implementation of RLHF (Reinforcement Learning with Human Feedback)
PaLM-rlhf-pytorch is a PyTorch implementation of Pathways Language Model (PaLM) with Reinforcement Learning from Human Feedback (RLHF). It is designed for fine-tuning large-scale language models with human preference alignment, similar to OpenAI’s approach for training models like ChatGPT. -
16
CLIP
CLIP, Predict the most relevant text snippet given an image
CLIP (Contrastive Language-Image Pretraining) is a neural model that links images and text in a shared embedding space, allowing zero-shot image classification, similarity search, and multimodal alignment. It was trained on large sets of (image, caption) pairs using a contrastive objective: images and their matching text are pulled together in embedding space, while mismatches are pushed apart. Once trained, you can give it any text labels and ask it to pick which label best matches a given image—even without explicit training for that classification task. ... -
17
YuE
Open source AI model for generating full songs from lyrics prompts
...It focuses on transforming text inputs such as lyrics and genre prompts into complete musical compositions that include both vocal and instrumental tracks. Unlike many shorter audio generators, the model is capable of producing songs that last several minutes while maintaining coherent musical structure and alignment with the provided lyrics. YuE introduces a family of models built on large language model architectures that process music generation as a sequence prediction task. YuE also incorporates techniques such as track-decoupled prediction and progressive conditioning to help manage complex audio signals and maintain consistency throughout long compositions. ... -
18
HivisionIDPhoto
HivisionIDPhotos: a lightweight and efficient AI ID photos tools
HivisionIDPhotos is an open-source AI project designed to automatically generate professional ID photographs from ordinary portrait images. The system uses computer vision and machine learning models to detect faces, segment the subject from the background, and produce standardized identification photos suitable for official documents. It is designed as a lightweight tool that can perform inference offline and run efficiently on CPUs without requiring powerful GPUs. The software analyzes... -
19
Tiktoken
tiktoken is a fast BPE tokeniser for use with OpenAI's models
tiktoken is a high-performance, tokenizer library (based on byte-pair encoding, BPE) designed for use with OpenAI’s models. It handles encoding and decoding text to token IDs efficiently, with minimal overhead. Because tokenization is a fundamental step in preparing text for models, tiktoken is optimized for speed, memory, and correctness in model contexts (e.g. matching OpenAI’s internal tokenization). The repo supports multiple encodings (e.g. “cl100k_base”) and lets users switch encoding... -
20
HunyuanVideo-Avatar
Tencent Hunyuan Multimodal diffusion transformer (MM-DiT) model
HunyuanVideo-Avatar is a multimodal diffusion transformer (MM-DiT) model by Tencent Hunyuan for animating static avatar images into dynamic, emotion-controllable, and multi-character dialogue videos, conditioned on audio. It addresses challenges of motion realism, identity consistency, and emotional alignment. Innovations include a character image injection module, an Audio Emotion Module for transferring emotion cues, and a Face-Aware Audio Adapter to isolate audio effects on faces, enabling multiple characters to be animated in a scene. Character image injection module for better consistency between training and inference conditioning. ... -
21
openTSNE
Extensible, parallel implementations of t-SNE
openTSNE is a modular Python implementation of t-Distributed Stochasitc Neighbor Embedding (t-SNE) [1], a popular dimensionality-reduction algorithm for visualizing high-dimensional data sets. openTSNE incorporates the latest improvements to the t-SNE algorithm, including the ability to add new data points to existing embeddings [2], massive speed improvements [3] [4] [5], enabling t-SNE to scale to millions of data points, and various tricks to improve the global alignment of the resulting visualizations. -
22
iFixAi
Independent Auditing of AI Agents
iFixAi is an independent auditing toolkit for evaluating whether an AI agent performs the business job it was assigned. Instead of measuring only latency, token use, or prompt-injection resistance, it examines operational behavior, organizational alignment, and task outcomes. A standard run performs 32 inspections across five evaluation pillars and produces an A-to-F grade. Users can configure and launch audits through a guided CLI, explicit command flags, or an agent plugin and skill. The same engine can test hosted providers or a real agent endpoint and can use self-grading, an independent judge, or a multi-judge ensemble. ... -
23
LLMSurvey
A Survey of Large Language Models
...The repository organizes hundreds of research papers into thematic sections that reflect the main areas of LLM research, including model architectures, training strategies, evaluation benchmarks, alignment techniques, and downstream applications. By structuring the literature in a navigable format, LLMSurvey allows researchers and practitioners to quickly explore important publications in the field without manually searching through multiple databases. -
24
Auto-Deep-Research
Your Fully-Automated Personal AI Assistant
...Auto-Deep-Research integrates retrieval from academic and web sources, processes document corpora for relevance and key insights, and organizes outputs into coherent chapters or sections according to research standards. It also embeds validation loops, where intermediate drafts are self-checked for consistency, coverage, and alignment with sound reasoning practices, reducing reliance on raw generation alone. -
25
Qwen2.5-Math
A series of math-specific large language models of our Qwen2 series
Qwen2.5-Math is a series of mathematics-specialized large language models in the Qwen2 family, released by Alibaba’s QwenLM. It includes base models (1.5B / 7B / 72B parameters), instruction-tuned versions, and a reward model (RM) to improve alignment. Unlike its predecessor Qwen2-Math, Qwen2.5-Math supports both Chain-of-Thought (CoT) reasoning and Tool-Integrated Reasoning (TIR) for solving math problems, and works in both Chinese and English. It is optimized for solving mathematical benchmarks and exams; the 72B-Instruct model achieves state-of-the-art results among open source models on many English and Chinese math tasks.
