Bridge Megatron-Core to Hugging Face/Reinforcement Learning
☆197Mar 2, 2026Updated this week
Alternatives and similar repositories for mbridge
Users that are interested in mbridge are comparing it to the libraries listed below
Sorting:
- (best/better) practices of megatron on veRL and tuning guide☆131Sep 26, 2025Updated 5 months ago
- Utility scripts for PyTorch (e.g. Make Perfetto show some disappearing kernels, Memory profiler that understands more low-level allocatio…☆90Sep 11, 2025Updated 5 months ago
- [Archived] For the latest updates and community contribution, please visit: https://github.com/Ascend/TransferQueue or https://gitcode.co…☆13Jan 16, 2026Updated last month
- Training library for Megatron-based models with bidirectional Hugging Face conversion capability☆481Updated this week
- Allow torch tensor memory to be released and resumed later☆220Feb 9, 2026Updated 3 weeks ago
- slime is an LLM post-training framework for RL Scaling.☆4,536Updated this week
- Best practices for training DeepSeek, Mixtral, Qwen and other MoE models using Megatron Core.☆167Jan 22, 2026Updated last month
- An external memory allocator example for PyTorch.☆16Aug 10, 2025Updated 6 months ago
- Ring attention implementation with flash attention☆987Sep 10, 2025Updated 5 months ago
- 🔥 LLM-powered GPU kernel synthesis: Train models to convert PyTorch ops into optimized Triton kernels via SFT+RL. Multi-turn compilation…☆127Nov 10, 2025Updated 3 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆185Feb 19, 2026Updated 2 weeks ago
- Implementation for FP8/INT8 Rollout for RL training without performence drop.☆293Nov 7, 2025Updated 4 months ago
- Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (…☆533Feb 26, 2026Updated last week
- The official repo of Pai-Megatron-Patch for LLM & VLM large scale training developed by Alibaba Cloud.☆1,534Dec 15, 2025Updated 2 months ago
- Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.☆934Updated this week
- ☆50Aug 21, 2025Updated 6 months ago
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference☆644Jan 15, 2026Updated last month
- A flexible and efficient training framework for large-scale alignment tasks☆451Oct 23, 2025Updated 4 months ago
- ☆21Mar 13, 2017Updated 8 years ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆716Updated this week
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆270Feb 2, 2026Updated last month
- A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training☆659Updated this week
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆469Feb 28, 2026Updated last week
- Pytorch Distributed native training library for LLMs/VLMs with OOTB Hugging Face support☆340Updated this week
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,264Aug 28, 2025Updated 6 months ago
- Efficient Long-context Language Model Training by Core Attention Disaggregation☆92Updated this week
- A scalable asynchronous reinforcement learning implementation with in-flight weight updates.☆373Feb 26, 2026Updated last week
- [Preprint] RLVE: Scaling Up Reinforcement Learning for Language Models with Adaptive Verifiable Environments☆184Jan 12, 2026Updated last month
- PSTensor provides a way to hack the memory management of tensors in TensorFlow and PyTorch by defining your own C++ Tensor Class.☆10Feb 10, 2022Updated 4 years ago
- Efficient retrieval head analysis with triton flash attention that supports topK probability☆13Jun 15, 2024Updated last year
- ☆12Mar 5, 2025Updated last year
- FPGA-based HyperLogLog Accelerator☆12Jul 13, 2020Updated 5 years ago
- ☆11Apr 3, 2023Updated 2 years ago
- A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.☆98Dec 17, 2025Updated 2 months ago
- A Benchmark for Multi-Stage Legal Case Documents Generation☆15Feb 24, 2025Updated last year
- VeOmni: Scaling Any Modality Model Training with Model-Centric Distributed Recipe Zoo☆1,676Updated this week
- Distributed parallel 3D-Causal-VAE for efficient training and inference☆47Aug 20, 2025Updated 6 months ago
- ☆129Jun 6, 2025Updated 9 months ago
- My learning notes for ML SYS.☆5,444Jan 30, 2026Updated last month