Skip to content

@CerebrasResearch CerebrasResearch

Change the repository type filter

All

    Repositories list

    10 repositories

    • cb_optillm

      Public
      Optimizing inference proxy for LLMs
      Python
      •
      Apache License 2.0
      •156•1•0•2•Updated Jan 23, 2025Jan 23, 2025

    • cb_short_transformers

      Public
      fork of short transformers with better dataset preparation and options
      Python
      •
      MIT License
      •0•0•0•0•Updated Dec 15, 2024Dec 15, 2024

    • nanoGNS

      Public
      Minimal reference implementations for per-example gradient norm methods for computing GNS
      Jupyter Notebook
      •1•6•0•0•Updated Nov 15, 2024Nov 15, 2024

    • LongBench

      Public
      LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding
      Python
      •
      MIT License
      •69•0•0•1•Updated Sep 3, 2024Sep 3, 2024

    • Cerebras-Trilabs

      Public
      Collaboration between Cerebras and DOE Tri-Labs (LLNL, LANL, and SNL)
      TeX
      •
      BSD 3-Clause "New" or "Revised" License
      •0•0•0•0•Updated Aug 6, 2024Aug 6, 2024

    • Sparse-IFT

      Public
      Official repository of Sparse ISO-FLOP Transformations for Maximizing Training Efficiency
      Python
      •
      Apache License 2.0
      •0•25•1•0•Updated Jul 31, 2024Jul 31, 2024

    • lm-evaluation-harness-master

      Public
      A framework for few-shot evaluation of language models.
      Python
      •
      MIT License
      •2k•0•0•0•Updated Jul 12, 2024Jul 12, 2024

    • specinfer-transformers

      Public
      🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
      Python
      •
      Apache License 2.0
      •28k•0•0•0•Updated Apr 26, 2024Apr 26, 2024

    • DeeperSpeed

      Public
      Cerebras's internal version of EleutherAI's DeeperSpeed library with bug fixes and patches made for our own version of GPT-Neox at https://github.com/CerebrasResearch/gpt-neox; DeepSpeed is a deep learning optimization library that makes distributed training easy, efficient, and effective.
      Python
      •
      Apache License 2.0
      •4.2k•0•0•0•Updated Oct 11, 2023Oct 11, 2023

    • RevBiFPN

      Public
      RevBiFPN: The Fully Reversible Bidirectional Feature Pyramid Network
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •1•12•0•0•Updated Oct 18, 2022Oct 18, 2022