Alternatives to Adapters

Open-source Python library for loading, fine-tuning, and running transformer models across text, vision, audio, and video.

RobBERT is a family of open pre-trained Dutch language models you fine-tune for classification, regression, and token-tagging tasks.

Ludwig is an open-source, YAML-driven deep learning framework built on PyTorch for training, fine-tuning, and serving multi-modal models.

MGM-Omni is an open-source omni-modal LLM demo fusing speech, text, and vision with zero-shot voice personalization.

Open-source multilingual BERT for 11 Indic languages and Indian-English.

Open-source, real-time end-to-end spoken dialogue model with voice cloning, self-hosted on your own infrastructure.

Falcon LLM delivers Apache-2.0 open-weight language and multimodal models you can self-host, from hybrid Transformer-Mamba to Arabic ASR

Open-weight coding model for 12-hour autonomous agent runs and 300-agent swarm orchestration.

Paper Instruments (formerly The LLM Data Company) ships open-weight frontier models for domain knowledge work, trained with on-policy reinforcement learning.

Magic builds frontier code models with ultra-long context to automate software engineering and AI research.

Linum trains text-to-video models from scratch and ships the weights under Apache 2.0 — download Linum v2, no render queue.

KoBigBird brings BigBird block sparse attention to Korean, handling up to 4096 tokens for long-document NLP.