align-anything
Align Anything: Training All-modality Model with Feedback
AI summary
Model aligner
Aligns large multimodal models with human intentions and values using various algorithms and fine-tuning methods.
- stars
- 270
- forks
- 53
- watching
- 9
Similar projects
Found by comparing what the projects do, not just their names.
Multimodal alignment model
Extending pretraining models to handle multiple modalities by aligning language and video representations
aidc-ai/ovis575
Multimodal aligner
An MLLM architecture designed to align visual and textual embeddings through structural alignment
Model alignment
This project demonstrates the effectiveness of reinforcement learning from human feedback (RLHF) in improving small language models like GPT-2.
Attention calibrator
This project proposes a novel method for calibrating attention distributions in multimodal models to improve contextualized representations of image-text pairs.
Model value alignment
Evaluates and aligns the values of Chinese large language models with safety and responsibility standards
Behavior alignment tool
Aligns large language models' behavior through fine-grained correctional human feedback to improve trustworthiness and accuracy.
Video benchmarking toolkit
Evaluates and benchmarks large language models' video understanding capabilities
Multimodal model evaluator
An open-source benchmark and evaluation tool for assessing multimodal large language models' performance in embodied decision-making tasks
Speech alignment tool
Scripts for aligning laboratory speech production data in Inuktitut
Mixture of Experts Model
A large vision-language model using a mixture-of-experts architecture to improve performance on multi-modal learning tasks
Speech alignment tool
Tools for aligning laboratory speech production data to forced audio alignment using HTK and SoX.
Multimodal model evaluator
Evaluating and improving large multimodal models through in-context learning
Multimodal benchmarking
Evaluates and benchmarks multimodal language models' ability to process visual, acoustic, and textual inputs simultaneously.
Language alignment tool
A Python library that uses machine learning and natural language processing to improve translation accuracy by aligning source and target languages
Model arena
An evaluation platform for comparing multi-modality models on visual question-answering tasks
