Table of Contents
1. First revision
2. Another Modification Request
3. Before we get too excited
Home Technology peripherals AI GPT-4o and SQL: How capable is a large model of changing its own architecture?

GPT-4o and SQL: How capable is a large model of changing its own architecture?

Jun 11, 2024 am 09:56 AM
sql Model GPT-4o

Author丨David Eastman

Compiled丨Noah

produced | 51CTO technology stack (WeChat ID: blog51cto)

Although there is no large language model ( LLM) have driven bicycles, but they clearly understand the role of driving behavior in the realm of human transportation. They are similar to what software developers provide as a kind of semantic-like real-world knowledge combined with an understanding of the technical world. We saw this clearly in a recent article where we were able to generate a simple SQL schema for book publishing simply by describing it in natural language.

While I was pleased with the performance of the Llama 3 creation schema, a colleague from my previous days at Oracle pointed out that the book publishing schema was a fairly familiar example. For ease of understanding, this is naturally a good thing, but in order to further expand the capabilities of LLM, in this article I will explore the ability of large language models to adjust their architecture according to problems described in English. This time, I will use OpenAI’s GPT-4o as it has served me well with code reviews recently.

As a starting point, we will start with the same questions as in the first article and summarize the answers. This answer is similar to last time. This time, GPT-4o not only provides us with an ERD (entity relationship diagram), but also explains the relationship between entities very well.

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

Similar to the previous attempt, it proposed the following architecture:

CREATE TABLE Author (  author_id INT AUTO_INCREMENT PRIMARY KEY,  first_name VARCHAR(50),  last_name VARCHAR(50),  birth_date DATE,  nationality VARCHAR(50) );  CREATE TABLE Publisher (  publisher_id INT AUTO_INCREMENT PRIMARY KEY,  name VARCHAR(100),  address VARCHAR(255),  contact_number VARCHAR(20),  email VARCHAR(100) );  CREATE TABLE Book (  book_id INT AUTO_INCREMENT PRIMARY KEY,  title VARCHAR(100),  genre VARCHAR(50),  publication_date DATE,  isbn VARCHAR(20) UNIQUE,  author_id INT,  publisher_id INT,  FOREIGN KEY (author_id) REFERENCES Author(author_id),  FOREIGN KEY (publisher_id) REFERENCES Publisher(publisher_id) );
Copy after login

I Prefer to have table names use the plural form of the contained objects, which I think is a widely accepted standard.

Large language models point out these relational limitations:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

So, using the same example data from last time, let’s Check whether the same result can be obtained in the SQL sandbox environment DB Fiddle.

If we populate this data and add the last view...

INSERT INTO Author (first_name, last_name, birth_date) VALUES ('Iain', 'Banks', '1954-02-16');  INSERT INTO Author (first_name, last_name, birth_date) VALUES ('Iain', 'M Banks', '1954-02-16');  INSERT INTO Publisher (name, address) VALUES ('Abacus', 'London');  INSERT INTO Publisher (name, address) VALUES ('Orbit', 'New York'); INSERT INTO Book (title, author_id, publisher_id, publication_date)VALUES ('Consider Phlebas', 2, 2, '1988-04-14'); INSERT INTO Book (title, author_id, publisher_id, publication_date)VALUES ('The Wasp Factory', 1, 1, '1984-02-15'); CREATE VIEW ViewableBooks ASSELECT Book.title 'Book', Author.first_name 'Author firstname', Author.last_name 'Author surname', Publisher.name 'Publisher', Book.publication_dateFROM Book, Publisher, AuthorWHERE Book.author_id = Author.author_idAND Book.publisher_id = Publisher.publisher_id;
Copy after login

We can get the desired result view from DB Fiddle in the table below:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

The second surname contains the middle name "M", which looks a bit awkward. Next, we will explore issues related to this.

1. First revision

As I mentioned in my previous article about SQL generation, "Ian Banks" and "Ian M Banks" are actually the same author. Last time, we didn't address this pen name issue. So, let's ask for the big model to fix this:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

So that's a good start. This time it needed to map the literary concept of "pen name" onto the existing architectural design it had produced. So it has to do more than just discover existing solutions. First, let's take a look at the newly established relationship:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

This seems reasonable. The following is the modified new table structure:

CREATE TABLE Pseudonym (  pseudonym_id INT AUTO_INCREMENT PRIMARY KEY,  pseudonym VARCHAR(100),  author_id INT,  FOREIGN KEY (author_id)  REFERENCES Author(author_id) );  CREATE TABLE Book (  book_id INT AUTO_INCREMENT PRIMARY KEY,  title VARCHAR(100),  genre VARCHAR(50),  publication_date DATE,  isbn VARCHAR(20) UNIQUE,  pseudonym_id INT,  publisher_id INT,  FOREIGN KEY (pseudonym_id) REFERENCES Pseudonym(pseudonym_id),  FOREIGN KEY (publisher_id) REFERENCES Publisher(publisher_id) );
Copy after login

This also feels right. The schema now associates books to pen names rather than directly to authors. Let's re-do a dbfiddle with the new schema, input the modified data to work with, and see if we can get the desired results again:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

Actually, the pen name column is just a field now, and the table looks neater.

2. Another Modification Request

Now I will request a further schema modification. We know that a book can have multiple authors (you may remember that Llama 3 proposed this without prompting last time), so we expect GPT-4o to revise its architecture again.

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

The new table that needs to be added is:

CREATE TABLE BookAuthor (  book_id INT,  pseudonym_id INT,  PRIMARY KEY (book_id, pseudonym_id),  FOREIGN KEY (book_id) REFERENCES Book(book_id),  FOREIGN KEY (pseudonym_id) REFERENCES Pseudonym(pseudonym_id) );
Copy after login

Thus, the relationship changes are as follows:

GPT-4o and SQL: How capable is a large model of changing its own architecture?Picture

(Note that there is a weird bracket error after describing the first few relationships. This The error is repeated in the description of all relationships. It seems to prevent the text "1:M" or "M:M" from being printed - maybe due to emoji confusion)

Of course, GPT- 4o is also following a single conversational thread – it takes its previous work into context. This much-lauded capability does make interacting with it more natural. Overall, it did a good job (and very quickly) of parsing our English descriptions to adapt its suggested schema.

3. Before we get too excited

Architecture is primarily about the relationships between things—it doesn’t require a deep understanding of the things themselves. However, this does not entirely mean that the road is clear for large models to take over database design.

Optimizing SQL queries and schemas has always been a bit of an art. You need to understand which common queries will be best suited for a certain design, how many tables will be involved, dependencies between queries, index definitions, partitioning, and so on. And that's before dealing with the CAP theorem dilemma - the trade-off between consistency and availability. Beneath these technical abstractions are expectations that data retrieval will be far more than simple.

I have no doubt that some combination of large language models and specialization will gradually solve these engineering problems over time, but for now we should be grateful for GPT-4o's ability to efficiently generate and modify reasonable architectures ability to feel victorious.

Reference link: https://thenewstack.io/gpt-4o-and-sql-how-well-can-an-llm-alter-its-own-schema/

To learn more about AIGC, please visit:

51CTO AI.x Community

https://www.51cto.com/aigc /

The above is the detailed content of GPT-4o and SQL: How capable is a large model of changing its own architecture?. For more information, please follow other related articles on the PHP Chinese website!

Statement of this Website
The content of this article is voluntarily contributed by netizens, and the copyright belongs to the original author. This site does not assume corresponding legal responsibility. If you find any content suspected of plagiarism or infringement, please contact admin@php.cn

Hot AI Tools

Undresser.AI Undress

Undresser.AI Undress

AI-powered app for creating realistic nude photos

AI Clothes Remover

AI Clothes Remover

Online AI tool for removing clothes from photos.

Undress AI Tool

Undress AI Tool

Undress images for free

Clothoff.io

Clothoff.io

AI clothes remover

Video Face Swap

Video Face Swap

Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Tools

Notepad++7.3.1

Notepad++7.3.1

Easy-to-use and free code editor

SublimeText3 Chinese version

SublimeText3 Chinese version

Chinese version, very easy to use

Zend Studio 13.0.1

Zend Studio 13.0.1

Powerful PHP integrated development environment

Dreamweaver CS6

Dreamweaver CS6

Visual web development tools

SublimeText3 Mac version

SublimeText3 Mac version

God-level code editing software (SublimeText3)

The world's most powerful open source MoE model is here, with Chinese capabilities comparable to GPT-4, and the price is only nearly one percent of GPT-4-Turbo The world's most powerful open source MoE model is here, with Chinese capabilities comparable to GPT-4, and the price is only nearly one percent of GPT-4-Turbo May 07, 2024 pm 04:13 PM

Imagine an artificial intelligence model that not only has the ability to surpass traditional computing, but also achieves more efficient performance at a lower cost. This is not science fiction, DeepSeek-V2[1], the world’s most powerful open source MoE model is here. DeepSeek-V2 is a powerful mixture of experts (MoE) language model with the characteristics of economical training and efficient inference. It consists of 236B parameters, 21B of which are used to activate each marker. Compared with DeepSeek67B, DeepSeek-V2 has stronger performance, while saving 42.5% of training costs, reducing KV cache by 93.3%, and increasing the maximum generation throughput to 5.76 times. DeepSeek is a company exploring general artificial intelligence

Hello, electric Atlas! Boston Dynamics robot comes back to life, 180-degree weird moves scare Musk Hello, electric Atlas! Boston Dynamics robot comes back to life, 180-degree weird moves scare Musk Apr 18, 2024 pm 07:58 PM

Boston Dynamics Atlas officially enters the era of electric robots! Yesterday, the hydraulic Atlas just "tearfully" withdrew from the stage of history. Today, Boston Dynamics announced that the electric Atlas is on the job. It seems that in the field of commercial humanoid robots, Boston Dynamics is determined to compete with Tesla. After the new video was released, it had already been viewed by more than one million people in just ten hours. The old people leave and new roles appear. This is a historical necessity. There is no doubt that this year is the explosive year of humanoid robots. Netizens commented: The advancement of robots has made this year's opening ceremony look like a human, and the degree of freedom is far greater than that of humans. But is this really not a horror movie? At the beginning of the video, Atlas is lying calmly on the ground, seemingly on his back. What follows is jaw-dropping

KAN, which replaces MLP, has been extended to convolution by open source projects KAN, which replaces MLP, has been extended to convolution by open source projects Jun 01, 2024 pm 10:03 PM

Earlier this month, researchers from MIT and other institutions proposed a very promising alternative to MLP - KAN. KAN outperforms MLP in terms of accuracy and interpretability. And it can outperform MLP running with a larger number of parameters with a very small number of parameters. For example, the authors stated that they used KAN to reproduce DeepMind's results with a smaller network and a higher degree of automation. Specifically, DeepMind's MLP has about 300,000 parameters, while KAN only has about 200 parameters. KAN has a strong mathematical foundation like MLP. MLP is based on the universal approximation theorem, while KAN is based on the Kolmogorov-Arnold representation theorem. As shown in the figure below, KAN has

Tesla robots work in factories, Musk: The degree of freedom of hands will reach 22 this year! Tesla robots work in factories, Musk: The degree of freedom of hands will reach 22 this year! May 06, 2024 pm 04:13 PM

The latest video of Tesla's robot Optimus is released, and it can already work in the factory. At normal speed, it sorts batteries (Tesla's 4680 batteries) like this: The official also released what it looks like at 20x speed - on a small "workstation", picking and picking and picking: This time it is released One of the highlights of the video is that Optimus completes this work in the factory, completely autonomously, without human intervention throughout the process. And from the perspective of Optimus, it can also pick up and place the crooked battery, focusing on automatic error correction: Regarding Optimus's hand, NVIDIA scientist Jim Fan gave a high evaluation: Optimus's hand is the world's five-fingered robot. One of the most dexterous. Its hands are not only tactile

FisheyeDetNet: the first target detection algorithm based on fisheye camera FisheyeDetNet: the first target detection algorithm based on fisheye camera Apr 26, 2024 am 11:37 AM

Target detection is a relatively mature problem in autonomous driving systems, among which pedestrian detection is one of the earliest algorithms to be deployed. Very comprehensive research has been carried out in most papers. However, distance perception using fisheye cameras for surround view is relatively less studied. Due to large radial distortion, standard bounding box representation is difficult to implement in fisheye cameras. To alleviate the above description, we explore extended bounding box, ellipse, and general polygon designs into polar/angular representations and define an instance segmentation mIOU metric to analyze these representations. The proposed model fisheyeDetNet with polygonal shape outperforms other models and simultaneously achieves 49.5% mAP on the Valeo fisheye camera dataset for autonomous driving

The latest from Oxford University! Mickey: 2D image matching in 3D SOTA! (CVPR\'24) The latest from Oxford University! Mickey: 2D image matching in 3D SOTA! (CVPR\'24) Apr 23, 2024 pm 01:20 PM

Project link written in front: https://nianticlabs.github.io/mickey/ Given two pictures, the camera pose between them can be estimated by establishing the correspondence between the pictures. Typically, these correspondences are 2D to 2D, and our estimated poses are scale-indeterminate. Some applications, such as instant augmented reality anytime, anywhere, require pose estimation of scale metrics, so they rely on external depth estimators to recover scale. This paper proposes MicKey, a keypoint matching process capable of predicting metric correspondences in 3D camera space. By learning 3D coordinate matching across images, we are able to infer metric relative

Single card running Llama 70B is faster than dual card, Microsoft forced FP6 into A100 | Open source Single card running Llama 70B is faster than dual card, Microsoft forced FP6 into A100 | Open source Apr 29, 2024 pm 04:55 PM

FP8 and lower floating point quantification precision are no longer the "patent" of H100! Lao Huang wanted everyone to use INT8/INT4, and the Microsoft DeepSpeed ​​team started running FP6 on A100 without official support from NVIDIA. Test results show that the new method TC-FPx's FP6 quantization on A100 is close to or occasionally faster than INT4, and has higher accuracy than the latter. On top of this, there is also end-to-end large model support, which has been open sourced and integrated into deep learning inference frameworks such as DeepSpeed. This result also has an immediate effect on accelerating large models - under this framework, using a single card to run Llama, the throughput is 2.65 times higher than that of dual cards. one

Comprehensively surpassing DPO: Chen Danqi's team proposed simple preference optimization SimPO, and also refined the strongest 8B open source model Comprehensively surpassing DPO: Chen Danqi's team proposed simple preference optimization SimPO, and also refined the strongest 8B open source model Jun 01, 2024 pm 04:41 PM

In order to align large language models (LLMs) with human values ​​and intentions, it is critical to learn human feedback to ensure that they are useful, honest, and harmless. In terms of aligning LLM, an effective method is reinforcement learning based on human feedback (RLHF). Although the results of the RLHF method are excellent, there are some optimization challenges involved. This involves training a reward model and then optimizing a policy model to maximize that reward. Recently, some researchers have explored simpler offline algorithms, one of which is direct preference optimization (DPO). DPO learns the policy model directly based on preference data by parameterizing the reward function in RLHF, thus eliminating the need for an explicit reward model. This method is simple and stable

See all articles