Preface
"Author Homepage":Sprite has white bubbles
"Personal Website" :Sprite’s personal website
2023 is the year of the explosion of large-scale artificial intelligence language models. Some concepts and English abbreviations also appear intensively in this year, which is easy to confuse and even confuse people.
Article directory
-
LLM: Large Language Model, a large language model, designed to understand and generate human language. LLM is characterized by its large scale and contains hundreds or hundreds of billions of parameters. It can capture the complex patterns of language, including syntax, semantics and some contextual information, thereby generating coherent and meaningful text. ChatGPT, GPT-4, BERT, Wen Xinyiyan, etc. are all typical large-scale language models.
-
GPT: Generative Pre-training Transformer, is a large-scale natural language generation model based on Transformer developed by OpenAI.
-
AIGC: Artificial Intelligence Generated Content, that is, AI-generated content. It refers to content generated using AI technology, such as AI writing articles, drawings and even videos, etc.
-
AGI: Artificial General Intelligence, that is, general artificial intelligence. The goal of AGI is to create a system that can think, learn, and perform a variety of tasks like humans, becoming an all-powerful "super brain" that may surpass humans in any field in the future.
In addition to concepts, if you want to learn more about the details and progress of these technologies, I recommend reading these books.
01 "ChatGPT driver software development"
Recommended words: Chinese IT leader Chen Bin’s new work explains in detail the application of ChatGPT in the entire software development process, greatly improving R&D efficiency and shaping engineers’ competitive advantages in the AI era.
import openai
# 设置OpenAI API密钥
openai.api_key = 'YOUR_API_KEY'
# 定义对话的起始信息
conversation_start = "User: Hello AI!\nAI: Hi, how can I help you today?"
# 发送请求并获取AI的回复
def get_ai_response(message):
response = openai.Completion.create(
engine='text-davinci-003',
prompt=conversation_start + message,
max_tokens=50,
temperature=0.7,
n = 1,
stop=None,
temperature = 0.6
)
return response.choices[0].text.strip()
# 与AI交互
while True:
user_input = input("User: ")
# 添加用户输入到对话中
conversation_start += '\nUser: ' + user_input
# 获取AI回复
ai_response = get_ai_response(conversation_start)
# 添加AI回复到对话中
conversation_start += '\nAI: ' + ai_response
print("AI:", ai_response)
02 "ChatGPT Principle and Practice"
Recommended words: Written by BAT senior AI experts and large model technology experts, highly recommended by MOSS system leader Qiu Xipeng and other experts! Systematically sort out and deeply analyze ChatGPT's core technology, algorithm implementation, working principles, and training methods, and provide a large amount of code and annotations.
03 "Neural Network and Deep Learning"
Recommended words: Douban score 9.5! The masterpiece of Professor Qiu Xipeng of Fudan University, recommended by Zhou Zhihua and Li Hang! The official version of the well-received Deep Learning Lectures Dandelion Book! Systematically organize the knowledge system of deep learning, and explain the principles, models and methods of deep learning from shallow to deep. A deep learning book more suitable for Chinese readers.
"Neural Networks and Deep Learning: Cases and Practice" is a supporting case of "Neural Networks and Deep Learning" produced by teacher Qiu Xipeng. It is deeply integrated with "Neural Networks and Deep Learning" to interpret the theoretical content of the original book from a practical perspective. Professor Qiu Xipeng of Fudan University and Baidu Flying Paddle R&D team jointly contributed.
import torch
import torch.nn as nn
import torch.optim as optim
import torchvision
import torchvision.transforms as transforms
# 定义神经网络模型
class NeuralNet(nn.Module):
def __init__(self):
super(NeuralNet, self).__init__()
self.fc1 = nn.Linear(784, 128) # 输入层到隐藏层
self.relu = nn.ReLU()
self.fc2 = nn.Linear(128, 10) # 隐藏层到输出层
def forward(self, x):
x = x.view(x.size(0), -1) # 将图像扁平化（将图像转换成一维向量）
x = self.fc1(x)
x = self.relu(x)
x = self.fc2(x)
return x
# 定义超参数
learning_rate = 0.001
batch_size = 100
num_epochs = 10
# 加载并预处理数据
transform = transforms.Compose([
transforms.ToTensor(),
transforms.Normalize((0.5,), (0.5,))
])
train_dataset = torchvision.datasets.MNIST(root='./data', train=True, transform=transform, download=True)
test_dataset = torchvision.datasets.MNIST(root='./data', train=False, transform=transform)
train_loader = torch.utils.data.DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
"AIGC Reshaping Education"
Recommendation:
Action Guide for Leading Education and Learning in the ChatGPT Era
Comprehensively help teachers, parents, and students stay ahead of the competition in the future Supporting video explanation, continuously updated Cutting edge knowledge in AIGC field Highly recommended by many educators and entrepreneurs
Written by Liu Wenyong, General Manager of College Student Business of Gaotu Education Technology Group
05 "General Artificial Intelligence"
Recommended words: A book on artificial intelligence that everyone can have. Since at least the 1950s, there has been much hype that a machine might soon be created that could match the full range and level of human intelligence. Now, we have successfully created machines that can solve specific problems with accuracy that matches or exceeds that of humans, but we still elude general intelligence. This book wants to discuss with you what kind of efforts are needed to obtain not only specialized intelligence, but also general intelligence.
If you're interested in intelligence, want to learn more about how to build autonomous machines, or are concerned that these machines will suddenly one day take over the world in something called the "technological singularity," read this book.