ModuleFormer is a MoE-based architecture that includes two different types of experts: stick-breaking attention heads and feedforward experts. We released a collection of ModuleFormer-based Language Models (MoLM) ranging in scale from 4 billion to 8 billion parameters.
☆223Sep 18, 2025Updated last year
Alternatives and similar repositories for ModuleFormer
Users that are interested in ModuleFormer are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Official repository for the paper "SwitchHead: Accelerating Transformers with Mixture-of-Experts Attention"☆102Sep 30, 2024Updated 2 years ago
- ☆415Nov 2, 2023Updated 2 years ago
- Triton-based implementation of Sparse Mixture of Experts.☆280Oct 3, 2025Updated last year
- ☆279Oct 31, 2023Updated 2 years ago
- Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".☆279Nov 3, 2023Updated 2 years ago
- GPU virtual machines on DigitalOcean Gradient AI • AdGet to production fast with high-performance AMD and NVIDIA GPUs you can spin up in seconds. The definition of operational simplicity.
- A family of open-sourced Mixture-of-Experts (MoE) Large Language Models☆1,702Mar 8, 2024Updated 2 years ago
- [ICLR 2024] Lemur: Open Foundation Models for Language Agents☆555Oct 28, 2023Updated 2 years ago
- [ACL 2023 Findings] Emergent Modularity in Pre-trained Transformers☆26Jun 7, 2023Updated 3 years ago
- Simplex Random Feature attention, in PyTorch☆76Oct 10, 2023Updated 3 years ago
- ☆16Dec 9, 2023Updated 2 years ago
- Reaching LLaMA2 Performance with 0.1M Dollars☆989Jul 23, 2024Updated 2 years ago
- ☆14Feb 1, 2024Updated 2 years ago
- some common Huggingface transformers in maximal update parametrization (µP)☆88Mar 14, 2022Updated 4 years ago
- Stick-breaking attention☆64Jul 1, 2025Updated last year
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- ☆30Sep 28, 2023Updated 3 years ago
- Fine-Tuning Pre-trained Transformers into Decaying Fast Weights☆20Oct 9, 2022Updated 4 years ago
- Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads☆2,775Jun 25, 2024Updated 2 years ago
- Code repository for the paper "MrT5: Dynamic Token Merging for Efficient Byte-level Language Models."☆59Sep 25, 2025Updated last year
- ☆130Jan 22, 2024Updated 2 years ago
- Repo for "Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture"☆564Dec 28, 2024Updated last year
- [ICLR 2023] "Sparse MoE as the New Dropout: Scaling Dense and Self-Slimmable Transformers" by Tianlong Chen*, Zhenyu Zhang*, Ajay Jaiswal…☆56Feb 28, 2023Updated 3 years ago
- Sequence Modeling with Multiresolution Convolutional Memory (ICML 2023)☆127Oct 11, 2023Updated 3 years ago
- ☆20May 30, 2024Updated 2 years ago
- AI Agents on DigitalOcean Gradient AI Platform • AdBuild production-ready AI agents using customizable tools or access multiple LLMs through a single endpoint. Create custom knowledge bases or connect external data.
- Repilot, a patch generation tool introduced in the ESEC/FSE'23 paper "Copiloting the Copilots: Fusing Large Language Models with Completi…☆138Oct 9, 2023Updated 3 years ago
- YaRN: Efficient Context Window Extension of Large Language Models☆1,803Apr 17, 2024Updated 2 years ago
- Positional Skip-wise Training for Efficient Context Window Extension of LLMs to Extremely Length (ICLR 2024)☆209May 20, 2024Updated 2 years ago
- [ICLR 2024] Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning☆643Mar 4, 2024Updated 2 years ago
- ☆1,597Mar 25, 2026Updated 6 months ago
- GRadient-INformed MoE☆264Sep 25, 2024Updated 2 years ago
- ☆93Aug 18, 2024Updated 2 years ago
- Inference code for Persimmon-8B☆411Sep 9, 2023Updated 3 years ago
- This is a new metric that can be used to evaluate faithfulness of text generated by LLMs. The work behind this repository can be found he…☆31Aug 25, 2023Updated 3 years ago
- Open source password manager - Proton Pass • AdSecurely store, share, and autofill your credentials with Proton Pass, the end-to-end encrypted password manager trusted by millions.
- Official PyTorch Implementation of EMoE: Unlocking Emergent Modularity in Large Language Models [main conference @ NAACL2024]☆40May 28, 2024Updated 2 years ago
- [ICML 2024] CLLMs: Consistency Large Language Models☆418Nov 16, 2024Updated last year
- Layer-Condensed KV cache w/ 10 times larger batch size, fewer params and less computation. Dramatic speed up with better task performance…☆155Apr 7, 2025Updated last year
- Triton Implementation of HyperAttention Algorithm☆48Dec 11, 2023Updated 2 years ago
- [ICLR 2025 & COLM 2025] Official PyTorch implementation of the Forgetting Transformer and Adaptive Computation Pruning☆154Feb 25, 2026Updated 7 months ago
- PB-LLM: Partially Binarized Large Language Models☆159Nov 20, 2023Updated 2 years ago
- Serving multiple LoRA finetuned LLM as one☆1,176May 8, 2024Updated 2 years ago