IST-DASLab / qmoeLinks
Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".
☆277Updated last year
Alternatives and similar repositories for qmoe
Users that are interested in qmoe are comparing it to the libraries listed below
Sorting:
- Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees"☆376Updated last year
- PB-LLM: Partially Binarized Large Language Models☆153Updated last year
- An efficent implementation of the method proposed in "The Era of 1-bit LLMs"☆154Updated 9 months ago
- ☆199Updated 8 months ago
- ☆549Updated 9 months ago
- scalable and robust tree-based speculative decoding algorithm☆351Updated 6 months ago
- Official PyTorch implementation of QA-LoRA☆138Updated last year
- Q-GaLore: Quantized GaLore with INT4 Projection and Layer-Adaptive Low-Rank Gradients.☆198Updated last year
- [ICML 2024] SqueezeLLM: Dense-and-Sparse Quantization☆697Updated 11 months ago
- Repository for the QUIK project, enabling the use of 4bit kernels for generative inference - EMNLP 2024☆181Updated last year
- Experiments on speculative sampling with Llama models☆128Updated 2 years ago
- GPTQ inference Triton kernel☆303Updated 2 years ago
- Advanced Ultra-Low Bitrate Compression Techniques for the LLaMA Family of LLMs☆110Updated last year
- Repo for "LoLCATs: On Low-Rank Linearizing of Large Language Models"☆244Updated 6 months ago
- ☆192Updated last week
- A high-throughput and memory-efficient inference and serving engine for LLMs☆265Updated 9 months ago
- For releasing code related to compression methods for transformers, accompanying our publications☆437Updated 6 months ago
- [ACL 2025 Main] EfficientQAT: Efficient Quantization-Aware Training for Large Language Models☆288Updated 2 months ago
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization☆365Updated 11 months ago
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"☆238Updated last month
- ModuleFormer is a MoE-based architecture that includes two different types of experts: stick-breaking attention heads and feedforward exp…☆223Updated last year
- ☆153Updated 2 years ago
- GPTQLoRA: Efficient Finetuning of Quantized LLMs with GPTQ☆103Updated 2 years ago
- Official code for "SWARM Parallelism: Training Large Models Can Be Surprisingly Communication-Efficient"☆141Updated last year
- Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clusters☆127Updated 8 months ago
- ☆546Updated 7 months ago
- [ICLR'25] Fast Inference of MoE Models with CPU-GPU Orchestration☆224Updated 8 months ago
- Code for paper: "QuIP: 2-Bit Quantization of Large Language Models With Guarantees" adapted for Llama models☆38Updated 2 years ago
- ☆127Updated last year
- ☆143Updated last month