跳到正文- Synthetic Data Generation
- Source: WizardLM / Evol-Instruct (Xu et al., 2023)
- Source: Will We Run Out of Data? (Villalobos et al., Epoch AI, 2024)
- Source: Vertex AI — Tune Gemini Models
- Source: verl Documentation
- Source: Unsloth Reinforcement Learning (GRPO) Guide
- Source: Unsloth — LoRA Hyperparameters Guide
- Source: Understanding R1-Zero-Like Training (Dr. GRPO)
- Source: Tulu 3 (Lambert et al., 2024) — RLVR
- Source: Training Language Models to Follow Instructions with Human Feedback (InstructGPT)
- Source: Training Compute-Optimal Large Language Models (Chinchilla, Hoffmann et al., 2022)
- Source: The Llama 3 Herd of Models (Meta, 2024)
- Source: The FineWeb Datasets (Penedo et al., Hugging Face, 2024)
- Source: The False Promise of Imitating Proprietary LLMs
- Source: The Curse of Recursion / Model Collapse (Shumailov et al.)
- Source: The Art of Scaling Reinforcement Learning Compute for LLMs (Khatri et al., 2025)
- Source: Textbooks Are All You Need (phi-1)
- Source: STaR — Self-Taught Reasoner (Zelikman et al., 2022)
- Source: Self-Instruct (Wang et al., 2022)
- Source: Scaling Laws for Neural Language Models (Kaplan et al., 2020)
- Source: Scaling Data-Constrained Language Models (Muennighoff et al., 2023)
- Source: RLAIF vs. RLHF
- Source: QLoRA: Efficient Finetuning of Quantized LLMs
- Source: OpenAI Reinforcement Fine-Tuning Guide
- Source: OpenAI Model Optimization Guide
- Source: OpenAI Deprecations — Fine-tuning Platform Wind-down
- Source: On-Policy Distillation of Language Models (GKD)
- Source: On-Policy Distillation (Thinking Machines)
- Source: Nemotron-4 340B Technical Report (NVIDIA, 2024)
- Source: Magpie (Xu et al., 2024)
- Source: LoRA: Low-Rank Adaptation of Large Language Models
- Source: LoRA Without Regret (Thinking Machines)
- Source: LoRA Learns Less and Forgets Less
- Source: LIMA: Less Is More for Alignment
- Source: Knowledge Distillation (Notebook)
- Source: Hugging Face TRL Documentation
- Source: Hugging Face PEFT — LoRA Developer Guide
- Source: Group Sequence Policy Optimization (GSPO)
- Source: Fine-tuning vs RAG vs Prompting (Notebook)
- Source: Fine-Tuning or Retrieval? Comparing Knowledge Injection in LLMs
- Source: Does RL Really Incentivize Reasoning Beyond the Base Model? (Yue et al., 2025)
- Source: Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations?
- Source: Distilling the Knowledge in a Neural Network
- Source: Direct Preference Optimization (DPO)
- Source: Deliberative Alignment (Guan et al., OpenAI, 2024)
- Source: DeepSeekMath and GRPO
- Source: DeepSeek-R1 (DeepSeek-AI, 2025)
- Source: DAPO (Yu et al., 2025)
- Source: Constitutional AI: Harmlessness from AI Feedback
- Source: Claude's new constitution (Anthropic, 2026)
- Source: Chinchilla Scaling: A Replication Attempt (Besiroglu et al., 2024)
- Source: Beyond Chinchilla-Optimal: Accounting for Inference in Scaling Laws (Sardana et al., 2024)
- Source: Axolotl Documentation
- Source: Are Emergent Abilities of Large Language Models a Mirage? (Schaeffer et al., 2023)
- Source: An Empirical Study of Catastrophic Forgetting in LLMs During Continual Fine-tuning
- Source: Amazon Bedrock — Model Customization
- Source: 2 OLMo 2 Furious (Team OLMo, Ai2, 2025)
- Scaling Laws
- Reinforcement Fine-Tuning
- Reasoning Models
- Preference Optimization
- PEFT and LoRA
- LLM Training Pipeline
- Knowledge Distillation
- Fine-tuning