
attention_sinks
by tomaarsen (personal account)
Extend existing LLMs way beyond the original training length with constant memory usage, without retraining (description from the project’s GitHub page)
Repository facts
- Stars
- 734
- Forks
- 43
- Open issues and pull requests
- 21
- Primary language
- Python
- License
- Apache License 2.0 (Apache-2.0)
- Owner
- tomaarsen · Personal account
- Created
- Oct 2, 2023
- Last push
- Apr 10, 2024
Topics
Related projects
Projects that share this one’s GitHub topics.
Rank 5,652: datadreamer-dev/DataDreamer
DataDreamer: Prompt. Generate Synthetic Data. Train & Align Models. 🤖💤
1,125 starsPythonMITLast push Feb 2, 2025
Rank 618: neuml/txtai
💡 All-in-one AI framework for semantic search, LLM orchestration and language model workflows
13,024 starsPythonApache-2.0Last push Oct 16, 2026
Rank 6,220: kennethleungty/Llama-2-Open-Source-LLM-CPU-Inference
Running Llama 2 and other Open-Source LLMs on CPU Inference Locally for Document Q&A
971 starsPythonMITLast push Nov 6, 2023
Rank 6,422: zamalali/DeepGit
Deep research agent to help you find the best GitHub repositories 🕵️!
922 starsPythonMITLast push Aug 30, 2026
Rank 6,577: explosion/curated-transformers
🤖 A PyTorch library of curated Transformer models and their composable components
890 starsPythonMITLast push Apr 17, 2024
Rank 8,119: webis-de/small-text
Active Learning for Text Classification in Python
647 starsPythonMITLast push Oct 12, 2026