Skip to content
Change the repository type filter

All

    Repositories list

    • EvoPress

      Public
      Python
      11700Updated Feb 13, 2025Feb 13, 2025
    • QuEST

      Public
      Work in progress.
      Jupyter Notebook
      MIT License
      13110Updated Feb 13, 2025Feb 13, 2025
    • PanzaMail

      Public
      Python
      Apache License 2.0
      1527855Updated Feb 10, 2025Feb 10, 2025
    • Official Repository for "Scalable Mechanistic Neural Networks" (ICLR 2025)
      MIT License
      0000Updated Feb 6, 2025Feb 6, 2025
    • gemm-fp8

      Public
      High Performance FP8 GEMM Kernels for SM89 and later GPUs.
      Cuda
      MIT License
      0300Updated Jan 24, 2025Jan 24, 2025
    • GridSearcher simplifies running grid searches for machine learning projects in Python, emphasizing parallel execution and GPU scheduling without dependencies on SLURM or other workload managers.
      Python
      Apache License 2.0
      0200Updated Jan 23, 2025Jan 23, 2025
    • gemm-int8

      Public
      High Performance Int8 GEMM Kernels for SM80 and later GPUs.
      Python
      MIT License
      0300Updated Jan 15, 2025Jan 15, 2025
    • MicroAdam

      Public
      This repository contains code for the MicroAdam paper.
      Python
      Apache License 2.0
      41610Updated Dec 14, 2024Dec 14, 2024
    • LLM training code for Databricks foundation models
      Python
      Apache License 2.0
      543001Updated Nov 27, 2024Nov 27, 2024
    • Python
      0000Updated Nov 25, 2024Nov 25, 2024
    • 0000Updated Nov 20, 2024Nov 20, 2024
    • LDAdam

      Public
      LDAdam - Adaptive Optimization from Low-Dimensional Gradient Statistics
      Python
      Apache License 2.0
      0600Updated Nov 6, 2024Nov 6, 2024
    • torch_cgx

      Public
      Pytorch distributed backend extension with compression support
      C++
      GNU Affero General Public License v3.0
      01640Updated Oct 17, 2024Oct 17, 2024
    • Python
      Apache License 2.0
      0700Updated Sep 5, 2024Sep 5, 2024
    • Boosting 4-bit inference kernels with 2:4 Sparsity
      Cuda
      Apache License 2.0
      46410Updated Sep 4, 2024Sep 4, 2024
    • marlin

      Public
      FP16xINT4 LLM inference kernel that can achieve near-ideal ~4x speedups up to medium batchsizes of 16-32 tokens.
      Python
      Apache License 2.0
      57705265Updated Sep 4, 2024Sep 4, 2024
    • sparsegpt

      Public
      Code for the ICML 2023 paper "SparseGPT: Massive Language Models Can Be Accurately Pruned in One-Shot".
      Python
      Apache License 2.0
      99763151Updated Aug 20, 2024Aug 20, 2024
    • peft-rosa

      Public
      A fork of the PEFT library, supporting Robust Adaptation (RoSA)
      Python
      Apache License 2.0
      31310Updated Aug 16, 2024Aug 16, 2024
    • Python
      MIT License
      0000Updated Jun 27, 2024Jun 27, 2024
    • spops

      Public
      C++
      Apache License 2.0
      0720Updated Jun 20, 2024Jun 20, 2024
    • Code for the EMNLP 2024 paper "Mathador-LM: A Dynamic Benchmark for Mathematical Reasoning on LLMs".
      Python
      Apache License 2.0
      0810Updated Jun 18, 2024Jun 18, 2024
    • SPADE

      Public
      Code of SPADE: Sparsity Guided Debugging for Deep Neural Networks
      Jupyter Notebook
      1110Updated May 25, 2024May 25, 2024
    • QUIK

      Public
      Repository for the QUIK project, enabling the use of 4bit kernels for generative inference - EMNLP 2024
      C++
      Apache License 2.0
      1417551Updated Apr 16, 2024Apr 16, 2024
    • FastOBQ-

      Public
      GPTQ with finetuning
      0000Updated Mar 27, 2024Mar 27, 2024
    • gptq

      Public
      Code for the ICLR 2023 paper "GPTQ: Accurate Post-training Quantization of Generative Pretrained Transformers".
      Python
      Apache License 2.0
      1642k231Updated Mar 27, 2024Mar 27, 2024
    • RoSA

      Public
      Official implementation of the ICML 2024 paper RoSA (Robust Adaptation)
      Python
      Apache License 2.0
      33810Updated Feb 13, 2024Feb 13, 2024
    • Repository for Sparse Finetuning of LLMs via modified version of the MosaicML llmfoundry
      Python
      Apache License 2.0
      64030Updated Jan 15, 2024Jan 15, 2024
    • CAP

      Public
      Repository for Correlation Aware Prune (NeurIPS23) source and experimental code
      Python
      Apache License 2.0
      1510Updated Nov 29, 2023Nov 29, 2023
    • qmoe

      Public
      Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".
      Python
      Apache License 2.0
      2226530Updated Nov 3, 2023Nov 3, 2023
    • ZipLM

      Public
      Code for the NeurIPS 2023 paper: "ZipLM: Inference-Aware Structured Pruning of Language Models".
      0210Updated Oct 20, 2023Oct 20, 2023