CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks
"LLM训练" 相关文章
Breaking Memory Limits: Gradient Wavelet Transform Enhances LLMs Training
Interviewing Ross Taylor on the state of AI: Chinese open models, scaling reasoning, useful tools, and what comes next
LLM Data Selection and Utilization via Dynamic Bi-level Optimization
Learning What Matters: Probabilistic Task Selection via Mutual Information for Model Finetuning
AI-Compass LLM训练框架生态：整合ms-swift、Unsloth、Megatron-LM等核心框架，涵盖全参数/PEFT训练与分布式优化
Scalpel vs. Hammer: GRPO Amplifies Existing Capabilities, SFT Replaces Them
WebSailor: Navigating Super-human Reasoning for Web Agent
Beyond First-Order: Training LLMs with Stochastic Conjugate Subgradients and AdamW
The Best Way to Align an LLM: Inner Alignment is Now a Solved Problem?
Revelo’s LatAm talent network sees strong demand from US companies, thanks to AI
历时6个月，Hugging Face开源LLM「超大规模实战手册」！200页3万字4000次训练
历时6个月，Hugging Face开源LLM「超大规模实战手册」！200页3万字4000次训练
Elevating AI Reasoning: The Art of Sampling for Learnability in LLM Training
LeCun力荐！进化算法淘汰77%低质数据：RIP方法让模型性能狂飙60%
LeCun力荐！进化算法淘汰77%低质数据：RIP方法让模型性能狂飙60%
Andrej Karpathy：我们需要让大模型“上学”，强化学习才刚开始
田渊栋：2024年年终总结
Frenzy: A Memory-Aware Serverless Computing Method for Heterogeneous GPU Clusters
Hugging Face Releases Picotron: A Tiny Framework that Solves LLM Training 4D Parallelization
