liangyuwang / Tiny-DeepSpeedLinks
Tiny-DeepSpeed, a minimalistic re-implementation of the DeepSpeed library
☆48Updated 4 months ago
Alternatives and similar repositories for Tiny-DeepSpeed
Users that are interested in Tiny-DeepSpeed are comparing it to the libraries listed below
Sorting:
- Efficient Mixture of Experts for LLM Paper List☆153Updated 2 months ago
- qwen-nsa☆85Updated 2 months ago
- ☆208Updated last month
- [ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length☆139Updated this week
- Implementation for FP8/INT8 Rollout for RL training without performence drop.☆281Updated last month
- 青稞Talk☆179Updated 2 weeks ago
- ☆126Updated 6 months ago
- ☆149Updated 5 months ago
- Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs☆198Updated 3 weeks ago
- MiroRL is an MCP-first reinforcement learning framework for deep research agent.☆184Updated 4 months ago
- DeepSeek Native Sparse Attention pytorch implementation☆110Updated last week
- Rethinking RL Scaling for Vision Language Models: A Transparent, From-Scratch Framework and Comprehensive Evaluation Scheme☆146Updated 8 months ago
- Super-Efficient RLHF Training of LLMs with Parameter Reallocation☆328Updated 8 months ago
- A lightweight reinforcement learning framework that integrates seamlessly into your codebase, empowering developers to focus on algorithm…☆93Updated 4 months ago
- Tiny-Megatron, a minimalistic re-implementation of the Megatron library☆19Updated 3 months ago
- A highly capable 2.4B lightweight LLM using only 1T pre-training data with all details.☆222Updated 5 months ago
- [ICLR 2025] COAT: Compressing Optimizer States and Activation for Memory-Efficient FP8 Training☆256Updated 4 months ago
- ☆65Updated last year
- ☆78Updated 3 weeks ago
- Bridge Megatron-Core to Hugging Face/Reinforcement Learning☆174Updated this week
- D^2-MoE: Delta Decompression for MoE-based LLMs Compression☆72Updated 9 months ago
- Exploring the Limit of Outcome Reward for Learning Mathematical Reasoning☆189Updated 9 months ago
- ☆87Updated 4 months ago
- Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (…☆450Updated this week
- TransMLA: Multi-Head Latent Attention Is All You Need (NeurIPS 2025 Spotlight)☆419Updated 3 months ago
- L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning☆258Updated 7 months ago
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆159Updated 2 months ago
- Spec-Bench: A Comprehensive Benchmark and Unified Evaluation Platform for Speculative Decoding (ACL 2024 Findings)☆344Updated 8 months ago
- Skywork-MoE: A Deep Dive into Training Techniques for Mixture-of-Experts Language Models☆138Updated last year
- ☆213Updated 10 months ago