Awesome LLM pre-training resources, including data, frameworks, and methods.
☆394Apr 29, 2025Updated last year
Alternatives and similar repositories for awesome-llm-pretraining
Users that are interested in awesome-llm-pretraining are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- A highly capable 2.4B lightweight LLM using only 1T pre-training data with all details.☆233Jul 25, 2025Updated last year
- ☆87Jul 21, 2026Updated last week
- slime is an LLM post-training framework for RL Scaling.☆7,679Updated this week
- LongAttn ：Selecting Long-context Training Data via Token-level Attention☆15Jul 16, 2025Updated last year
- An Easy-to-use, Scalable and High-performance Agentic RL Framework based on Ray (PPO & DAPO & REINFORCE++ & VLM & TIS & vLLM & Ray & Asy…☆9,855Jul 14, 2026Updated 2 weeks ago
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- A series of technical report on Slow Thinking with LLM☆767Aug 13, 2025Updated 11 months ago
- Collection of papers for scalable automated alignment.☆92Oct 22, 2024Updated last year
- verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework☆22,699Updated this week
- ☆20Jun 27, 2026Updated last month
- Minimalistic large language model 3D-parallelism training☆2,768May 26, 2026Updated 2 months ago
- A Survey of Reinforcement Learning for Large Reasoning Models☆2,470Nov 9, 2025Updated 8 months ago
- Simple and scalable tools for data-driven pretraining data selection.☆30Jun 9, 2025Updated last year
- The OlymMATH dataset☆24Jun 1, 2025Updated last year
- 🔥 A minimal training framework for scaling FLA models☆406Apr 22, 2026Updated 3 months ago
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- ☆71Oct 16, 2024Updated last year
- The official repository for SkyLadder: Better and Faster Pretraining via Context Window Scheduling☆43Dec 29, 2025Updated 7 months ago
- Ongoing research training transformer models at scale☆17,247Updated this week
- The loss landscape of Large Language Models resemble basin!☆41Jul 8, 2025Updated last year
- An Efficient and User-Friendly Scaling Library for Reinforcement Learning with Large Language Models☆3,333Updated this week
- Spectral Sphere Optimizer☆131Mar 23, 2026Updated 4 months ago
- Pretrain、decay、SFT a CodeLLM from scratch 🧙♂️☆40May 15, 2024Updated 2 years ago
- Awesome Reasoning LLM Tutorial/Survey/Guide☆2,493Updated this week
- Codes and Data for Scaling Relationship on Learning Mathematical Reasoning with Large Language Models☆270Sep 12, 2024Updated last year
- Serverless GPU API endpoints on Runpod - Get Bonus Credits • AdSkip the infrastructure headaches. Auto-scaling, pay-as-you-go, no-ops approach lets you focus on innovating your application.
- A framework for few-shot evaluation of language models.☆13,443Jul 13, 2026Updated 2 weeks ago
- An efficient implementation of the NSA (Native Sparse Attention) kernel☆133Jun 24, 2025Updated last year
- 🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"☆1,014Feb 5, 2026Updated 5 months ago
- Official Repo for Open-Reasoner-Zero☆2,096Jun 2, 2025Updated last year
- Physics of Language Models: Part 4.2, Canon Layers at Scale where Synthetic Pretraining Resonates in Reality☆356May 20, 2026Updated 2 months ago
- From Chain-of-Thought prompting to OpenAI o1 and DeepSeek-R1 🍓☆3,657Apr 20, 2026Updated 3 months ago
- ☆155Jul 4, 2025Updated last year
- 🚀 Efficient implementations for emerging model architectures☆5,463Updated this week
- Based on the R1-Zero method, using rule-based rewards and GRPO on the Code Contests dataset.☆18Apr 22, 2025Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- Reproduce R1 Zero on Logic Puzzle☆2,450Mar 20, 2025Updated last year
- ☆25Dec 13, 2024Updated last year
- A lightweight, AI-native training framework for large language models. Designed for fast iteration, reproducible experiments, and modular…☆579May 18, 2026Updated 2 months ago
- [COLM 2025] An Open Math Pre-trainng Dataset with 370B Tokens.☆110Apr 4, 2025Updated last year
- R1-searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning☆720Aug 5, 2025Updated 11 months ago
- Use the tokenizer in parallel to achieve superior acceleration☆20Mar 21, 2024Updated 2 years ago
- Parallel Scaling Law for Language Model — Beyond Parameter and Inference Time Scaling☆480May 17, 2025Updated last year