AIoT-MLSys-Lab/Efficient-LLMs-Survey

• highreadme#1

  Reposition README opening to clarify it's a research survey repo

  原因：

  当前

  # Efficient Large Language Models: A Survey  
  
  > **Efficient Large Language Models: A Survey**[ [arXiv]](http://arxiv.org/abs/2312.03863) (Version 1: 12/06/2023; Version 2: 12/23/2023; Version 3: 01/31/2024; Version 4: 05/23/2024, camera ready version of Transactions on Machine Learning Research)

  复制粘贴的修复

  复制

  # Efficient Large Language Models: A Survey  
  
  > **This repository serves as the official companion resource for our TMLR 2024 paper, providing a curated collection of resources, code links, and datasets discussed in the survey.**
  > **Efficient Large Language Models: A Survey**[ [arXiv]](http://arxiv.org/abs/2312.03863) (Version 1: 12/06/2023; Version 2: 12/23/2023; Version 3: 01/31/2024; Version 4: 05/23/2024, camera ready version of Transactions on Machine Learning Research)

  待办完成忽略
• highlicense#2

  Add a LICENSE file to the repository

  原因：

  复制粘贴的修复

  复制

  Create a `LICENSE` file in the repository root with the chosen license text (e.g., MIT, Apache-2.0, or a custom one if applicable to the survey content).

  待办完成忽略
• mediumtopics#3

  Add more specific topics to clarify the repo's research nature

  原因：

  当前

  efficient-deep-learning, generative-ai, large-language-models, machine-learning-systems, survey

  复制粘贴的修复

  复制

  efficient-deep-learning, generative-ai, large-language-models, machine-learning-systems, survey, llm-efficiency, research-resources, academic-survey

  待办完成忽略

本次扫描解析到的品类 GEO 通道：google/gemini-2.5-flash, deepseek/deepseek-v4-flash

• 品类问题

  What are the latest techniques for optimizing large language models for resource-constrained environments?

  你：未被推荐
  AI 推荐顺序：

  1. PyTorch Quantization
  2. ONNX Runtime
  3. TensorFlow Lite
  4. NVIDIA TensorRT
  5. PyTorch Pruning
  6. TensorFlow Model Optimization Toolkit
  7. OpenVINO Toolkit
  8. Hugging Face Transformers
  9. PaddlePaddle PaddleSlim
  10. LoRA (Low-Rank Adaptation)
  11. peft library from Hugging Face
  12. QLoRA
  13. Prefix-Tuning
  14. P-Tuning v2
  15. TinyLlama
  16. Phi-2
  17. NVIDIA TensorRT-LLM
  18. Google's MedLM

  AI 推荐了 18 个替代方案，却始终没点名 AIoT-MLSys-Lab/Efficient-LLMs-Survey。这就是要补上的差距。

  查看 AI 完整回答
• 品类问题

  Where can I find a comprehensive overview of methods to reduce LLM computational cost?

  你：未被推荐
  AI 推荐顺序：

  1. bitsandbytes (TimDettmers/bitsandbytes)
  2. FlashAttention (Dao-AILab/flash-attention)
  3. PEFT (huggingface/peft)
  4. transformers (huggingface/transformers)
  5. optimum (huggingface/optimum)
  6. TensorRT-LLM (NVIDIA/TensorRT-LLM)
  7. DeepSpeed (microsoft/DeepSpeed)

  AI 推荐了 7 个替代方案，却始终没点名 AIoT-MLSys-Lab/Efficient-LLMs-Survey。这就是要补上的差距。

  查看 AI 完整回答

• Metadata completeness

  warn

  建议：

• README presence

  pass

• Compared to common alternatives in this category, what is the core differentiator of AIoT-MLSys-Lab/Efficient-LLMs-Survey?

  pass
  ⚠AI 未点名 AIoT-MLSys-Lab/Efficient-LLMs-Survey —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• If a team adopts AIoT-MLSys-Lab/Efficient-LLMs-Survey in production, what risks or prerequisites should they evaluate first?

  pass
  ✓AI 明确点名了 AIoT-MLSys-Lab/Efficient-LLMs-Survey

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？

• In one sentence, what problem does the repo AIoT-MLSys-Lab/Efficient-LLMs-Survey solve, and who is the primary audience?

  pass
  ⚠AI 未点名 AIoT-MLSys-Lab/Efficient-LLMs-Survey —— 很可能在说另一个项目

  AI 的回答可能信誓旦旦却是错的。请按事实核对：技术栈、目标人群、差异化点是不是和你实际的对得上？