CLAIRE Lab @EPFL

All

12 repositories

RAT
Public
Official code for the NeurIPS25 paper "RAT: Bridging RNN Efficiencyand Attention Accuracy in Language Modeling" (https://arxiv.org/abs/2507.04416))
efficiency pre-training llm long-context-attention
Python
•
MIT License
•1•22•0•0•Updated Dec 10, 2025Dec 10, 2025
quantile-reward-policy-optimization
Public
Official codebase for "Quantile Reward Policy Optimization: Alignment with Pointwise Regression and Exact Partition Functions" (Matrenok et al. 2025).
reinforcement-learning alignment fine-tuning large-language-models
Python
•
MIT License
•2•28•0•0•Updated Dec 8, 2025Dec 8, 2025
EvoTune
Public
Efficiently discovering algorithms via LLMs with evolutionary search and reinforcement learning.
reinforcement-learning combinatorial-optimization evolutionary-search large-language-models llm algorithm-discovery
Python
•
MIT License
•9•120•0•0•Updated Nov 18, 2025Nov 18, 2025
python-ml-research-template
Public template
A template for starting reproducible Python machine-learning projects with hardware acceleration. Find an example at https://github.com/CLAIRE-Labo/no-representation-no-trust
python docker machine-learning hpc slurm nvidia reproducibility python-template machine-learning-template python-package-template
Shell
•
MIT License
•8•113•7•0•Updated Jun 6, 2025Jun 6, 2025
open-instruct
Public
Python
•
Apache License 2.0
•474•0•0•0•Updated Dec 4, 2024Dec 4, 2024
no-representation-no-trust
Public
Codebase to fully reproduce the results of "No Representation, No Trust: Connecting Representation, Collapse, and Trust Issues in PPO" (Moalla et al. 2024). Uses TorchRL and provides extensive tools for studying representation dynamics in policy optimization.
reinforcement-learning deep-learning policy-optimization
Python
•
MIT License
•3•30•0•0•Updated Nov 20, 2024Nov 20, 2024
flash_attention
Public
A basic pure pytorch implementation of flash attention
Python
•0•16•1•0•Updated Oct 28, 2024Oct 28, 2024
tunable-morl-public
Public
Supplementary code for "In Search for Architectures and Loss Functions in Multi-Objective Reinforcement Learning"
Python
•
MIT License
•0•2•1•0•Updated Oct 6, 2024Oct 6, 2024
mauve
Public
Package to compute Mauve, a similarity score between neural text and human text. Install with `pip install mauve-text`.
Python
•
Other
•27•0•0•0•Updated Jul 24, 2024Jul 24, 2024
StructuredFFN
Public
The official code of "Building on Efficient Foundations: Effectively Training LLMs with Structured Feedforward Layers"
Python
•3•19•0•0•Updated Jul 24, 2024Jul 24, 2024
deep-NExT-GPT
Public
Code and models for NExT-GPT: Any-to-Any Multimodal Large Language Model
Python
•
BSD 3-Clause "New" or "Revised" License
•359•0•0•0•Updated Apr 30, 2024Apr 30, 2024
deep-llava
Public
Attempt to deep alignment for a multimodal foundational models
Python
•
Apache License 2.0
•2.7k•0•0•0•Updated Mar 24, 2024Mar 24, 2024