Table of Contents
Open-Platypus Dataset
Contamination problem
Fine-tuning and merging
The results
Home Technology peripherals AI The Open LLM list has been refreshed again, and a 'Platypus' stronger than Llama 2 is here.

The Open LLM list has been refreshed again, and a 'Platypus' stronger than Llama 2 is here.

Aug 17, 2023 pm 03:09 PM
ai Model

In order to challenge the dominance of closed models such as OpenAI’s GPT-3.5 and GPT-4, a series of open source models are emerging, including LLaMa, Falcon, etc. Recently, Meta AI launched LLaMa-2, which is known as the most powerful model in the open source field, and many researchers have also built their own models on this basis. For example, StabilityAI used Orca-style data sets to fine-tune the Llama2 70B model and developed StableBeluga2, which also achieved good results on Huggingface's Open LLM rankings

The latest Open The LLM list ranking has changed, and the Platypus (Platypus) model has successfully climbed to the top of the list

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

The author is from Boston University and uses PEFT and LoRA And the dataset Open-Platypus fine-tuned and optimized Platypus based on Llama 2

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

The author introduced Platypus in detail in a paper

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

The paper can be found at: https://arxiv.org/abs/2308.07317

The following are the main contributions of this article:

  • Open-Platypus is a small-scale dataset consisting of a curated subset of public text datasets . This dataset consists of 11 open source datasets with a focus on improving LLM’s STEM and logic knowledge. It consists mainly of questions designed by humans, with only 10% of questions generated by LLM. The main advantage of Open-Platypus is its scale and quality, which enables very high performance in a short time and with low time and cost of fine-tuning. Specifically, training a 13B model using 25k problems takes just 5 hours on a single A100 GPU.
  • Describes the similarity elimination process, reduces the size of the dataset, and reduces data redundancy.
  • The ever-present phenomenon of contamination of open LLM training sets with data contained in important LLM test sets is analyzed in detail, and the author's training data filtering process to avoid this hidden danger is introduced.
  • Describes the process of selecting and merging specialized fine-tuned LoRA modules.

Open-Platypus Dataset

The author has currently released the Open-Platypus Dataset on Hugging Face


Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

Contamination problem

To avoid benchmarking problems leaking into the training set, This approach first considers preventing this problem to ensure that the results are not simply biased by memory. While striving for accuracy, the authors are also aware of the need for flexibility in marking please say again questions because questions can be asked in a variety of ways and are influenced by general domain knowledge. To manage potential leakage issues, the authors carefully designed heuristics for manually filtering problems with more than 80% similarity to the cosine embedding of the benchmark problem in Open-Platypus. They divided potential leak issues into three categories: (1) Please say the question again; (2) Rephrase: This area presents a gray toned problem; (3) similar but not identical problem. To be cautious, they excluded all of these problems from the training set

Please say it again

This text almost exactly replicates the content of the test question set, with only slight modifications or rearrangements of the words. Based on the number of leaks in the table above, the authors believe this is the only category that falls under contamination. The following are specific examples:

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

Redescription: This area has a gray tint

The following issues are called redescriptions: This area takes on a shade of gray and includes issues that are not exactly, please, common sense. While the authors leave the final judgment on these issues to the open source community, they argue that these issues often require expert knowledge. It should be noted that this type of questions includes questions with exactly the same instructions but synonymous answers:

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

Similar but not identical

These questions have a high degree of similarity, but due to subtle changes between the questions, there are significant differences in the answers.

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

Fine-tuning and merging

After the data set is improved, the author focuses on two methods: low Rank approximation (LoRA) training and parameter efficient fine-tuning (PEFT) library. Unlike full fine-tuning, LoRA retains the weights of the pre-trained model and uses the rank decomposition matrix for integration in the transformer layer, thereby reducing trainable parameters and saving training time and cost. Initially, fine-tuning mainly focused on attention modules such as v_proj, q_proj, k_proj and o_proj. Subsequently, it was extended to the gate_proj, down_proj and up_proj modules according to the suggestions of He et al. Unless the trainable parameters are less than 0.1% of the total parameters, these modules all show better results. The author adopted this method for both the 13B and 70B models, and the result was that the trainable parameters were 0.27% and 0.2% respectively. The only difference is the initial learning rate of these models

The results

According to the Hugging Face Open LLM ranking data on August 10, 2023, The author compared Platypus with other SOTA models and found that the Platypus2-70Binstruct variant performed well, ranking first with an average score of 73.13

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

Stable -Platypus2-13B model stands out with an average score of 63.96 among 13 billion parameter models, which deserves attention

Open LLM榜单再次刷新,比Llama 2更强的「鸭嘴兽」来了

##Limitations

Platypus, as a fine-tuned extension of LLaMa-2, retains many of the constraints of the base model and introduces specific challenges through targeted training. It shares the static knowledge base of LLaMa-2, which may become outdated . Additionally, there is a risk of generating inaccurate or inappropriate content, particularly in cases of unclear prompts. While Platypus has been enhanced in STEM and English logic, its proficiency in other languages ​​is not reliable and may be inconsistent. It occasionally produces biased or inconsistent harmful content. The author acknowledges efforts to minimize these issues but acknowledges the ongoing challenges, particularly in non-English languages.

The potential for abuse of Platypus is a concern. issues, so developers should conduct security testing of their applications before deployment. Platypus may have some limitations outside of its primary domain, so users should proceed with caution and consider additional fine-tuning for optimal performance. Users need to ensure that the training data for Platypus does not overlap with other benchmark test sets. The authors are very cautious about data contamination issues and avoid merging models with models trained on tainted datasets. Although it is confirmed that there is no contamination in the cleaned training data, it cannot be ruled out that some problems may have been overlooked. For details on these limitations, see the Limitations section in the paper

The above is the detailed content of The Open LLM list has been refreshed again, and a 'Platypus' stronger than Llama 2 is here.. 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

AI Hentai Generator

AI Hentai Generator

Generate AI Hentai for free.

Hot Article

R.E.P.O. Energy Crystals Explained and What They Do (Yellow Crystal)
2 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
R.E.P.O. Best Graphic Settings
2 weeks ago By 尊渡假赌尊渡假赌尊渡假赌
R.E.P.O. How to Fix Audio if You Can't Hear Anyone
2 weeks ago By 尊渡假赌尊渡假赌尊渡假赌

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)

How to play picture sequences smoothly with CSS animation? How to play picture sequences smoothly with CSS animation? Apr 04, 2025 pm 05:57 PM

How to achieve the playback of pictures like videos? Many times, we need to implement similar video player functions, but the playback content is a sequence of images. direct...

Zustand asynchronous operation: How to ensure the latest state obtained by useStore? Zustand asynchronous operation: How to ensure the latest state obtained by useStore? Apr 04, 2025 pm 02:09 PM

Data update problems in zustand asynchronous operations. When using the zustand state management library, you often encounter the problem of data updates that cause asynchronous operations to be untimely. �...

How to quickly build a foreground page in a React Vite project using AI tools? How to quickly build a foreground page in a React Vite project using AI tools? Apr 04, 2025 pm 01:45 PM

How to quickly build a front-end page in back-end development? As a backend developer with three or four years of experience, he has mastered the basic JavaScript, CSS and HTML...

How to implement nesting effect of text annotations in Quill editor? How to implement nesting effect of text annotations in Quill editor? Apr 04, 2025 pm 05:21 PM

A solution to implement text annotation nesting in Quill Editor. When using Quill Editor for text annotation, we often need to use the Quill Editor to...

How to achieve the effect of high input elements but high text at the bottom? How to achieve the effect of high input elements but high text at the bottom? Apr 04, 2025 pm 10:27 PM

How to achieve the height of the input element is very high but the text is located at the bottom. In front-end development, you often encounter some style adjustment requirements, such as setting a height...

How to solve the problem that the result of OpenCV.js projection transformation is a blank transparent picture? How to solve the problem that the result of OpenCV.js projection transformation is a blank transparent picture? Apr 04, 2025 pm 03:45 PM

How to solve the problem of transparent image with blank projection transformation result in OpenCV.js. When using OpenCV.js for image processing, sometimes you will encounter the image after projection transformation...

How to use CSS to achieve smooth playback effect of image sequences? How to use CSS to achieve smooth playback effect of image sequences? Apr 04, 2025 pm 04:57 PM

How to realize the function of playing pictures like videos? Many times, we need to achieve similar video playback effects in the application, but the playback content is not...

How to implement notifications before task start using Quartz timer and cron expression without changing the front end? How to implement notifications before task start using Quartz timer and cron expression without changing the front end? Apr 04, 2025 pm 02:15 PM

Implementation method of task scheduling notification In task scheduling, the Quartz timer uses cron expression to determine the execution time of the task. Now we are facing this...

See all articles