NielsRogge/Transformers-Tutorials

Covers multimodal and vision tasks (audio classification, object detection, semantic segmentation, depth estimation) alongside NLP applications through PyTorch-based tutorials. Demonstrates architecture-specific techniques like fine-tuning `BertForTokenClassification` for NER, zero-shot `CLIPSeg` inference, and DETR variants for detection and panoptic segmentation. Integrates HuggingFace ecosystem components including Transformers, Tokenizers, Datasets, and Accelerate with Colab notebooks for reproducible experimentation.

11,519 stars. Actively maintained with 5 commits in the last 30 days.