epfml / landmark-attentionLinks
Landmark Attention: Random-Access Infinite Context Length for Transformers
☆425Updated last year
Alternatives and similar repositories for landmark-attention
Users that are interested in landmark-attention are comparing it to the libraries listed below
Sorting:
- ☆416Updated last year
- Multipack distributed sampler for fast padding-free training of LLMs☆201Updated last year
- A bagel, with everything.☆325Updated last year
- batched loras☆346Updated 2 years ago
- The Truth Is In There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction☆388Updated last year
- [ICML'24 Spotlight] LLM Maybe LongLM: Self-Extend LLM Context Window Without Tuning☆661Updated last year
- Merge Transformers language models by use of gradient parameters.☆208Updated last year
- ☆96Updated 2 years ago
- Official repository for LongChat and LongEval☆533Updated last year
- ☆459Updated last year
- [ICML 2024] SqueezeLLM: Dense-and-Sparse Quantization☆703Updated last year
- Extend existing LLMs way beyond the original training length with constant memory usage, without retraining☆721Updated last year
- Official code for ReLoRA from the paper Stack More Layers Differently: High-Rank Training Through Low-Rank Updates☆463Updated last year
- Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".☆277Updated last year
- ☆546Updated 9 months ago
- Code for the paper "Rethinking Benchmark and Contamination for Language Models with Rephrased Samples"☆310Updated last year
- Load multiple LoRA modules simultaneously and automatically switch the appropriate combination of LoRA modules to generate the best answe…☆158Updated last year
- Inference code for Mistral and Mixtral hacked up into original Llama implementation☆370Updated last year
- Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees"☆383Updated last year
- This is our own implementation of 'Layer Selective Rank Reduction'☆240Updated last year
- ☆275Updated 2 years ago
- Positional Skip-wise Training for Efficient Context Window Extension of LLMs to Extremely Length (ICLR 2024)☆206Updated last year
- ☆570Updated last year
- TART: A plug-and-play Transformer module for task-agnostic reasoning☆201Updated 2 years ago
- ModuleFormer is a MoE-based architecture that includes two different types of experts: stick-breaking attention heads and feedforward exp…☆224Updated 2 weeks ago
- ☆371Updated 2 years ago
- Experiments on speculative sampling with Llama models☆128Updated 2 years ago
- Inference code for Persimmon-8B☆414Updated 2 years ago
- Official PyTorch implementation of QA-LoRA☆141Updated last year
- Spherical Merge Pytorch/HF format Language Models with minimal feature loss.☆138Updated 2 years ago