人工智能

catalog

concept

abbr	full	description
NLP	Natural Language Processing	自然语言处理
LLMs	Large Language Models	大语言模型
GLM	General Language Model	通用语言模型
LoRA	Low-Rank Adaptation of Large Language Models	大语言模型的低阶适应
BERT	Bidirectional Encoder Representations from Transformers	基于大规模无监督预训练的双向表示法
GPT	Generative Pre-trained Transformer	生成式预训练语言模型
LLaMA	Large Language Model Meta AI	facebook 大语言模型
RNN	Recurrent Neural Network	循环神经网络
CNN	Convolutional Neural Network	卷积神经网络
ELMO	Embeddings from Language Models
SA	Self-Attention	自注意力机制
F-FNN	Feed-Forward Neural Networks	前向神经网络
RC	Residual Connection	残差连接
MLM	Masked Language Model	掩码语言模型
NSP	Next Sentence Prediction	下一句预测
POS	Part of Speech	词性
TF	Term Frequency	词频
SW	Stop Words	停用词
IDF	Inverse Document Frequency	逆文档频率
FT	Fine Tuning	微调

TensorFlow vs PyTorch vs Jax

• TensorFlow

  An Open Source Machine Learning Framework for Everyone

• PyTorch

  Tensors and Dynamic neural networks in Python with strong GPU acceleration

• Jax

  Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Transformer vs RNN vs CNN

• Transformer : (Multi-Head Attention) 多头自注意力机制

  特色 : 自然语言处理

• RNN : 循环神经网络

  特色 : 处理序列数据和时序关系

• CNN : 卷积神经网络

  特色 : 处理网格数据(如: 图像处理)