Maxime Labonne's picture

Maxime Labonne PRO

mlabonne

·

https://mlabonne.github.io/blog

AI & ML interests

Post-training, model editing, quantization

Recent Activity

upvoted a paper about 18 hours ago

Enhancing Human-Like Responses in Large Language Models

replied to CultriX's post about 22 hours ago

# Space for Multi-Agent Workflows using AutoGen Hi all, I created this "AutoGen Multi-Agent Workflow" space that allows you to experiment with multi-agent workflows. By default, it allows code generation with built-in quality control and automatic documentation generation. It achieves this by leveraging multiple AI agents working together to produce high-quality code snippets, ensuring they meet the specified requirements. In addition to the default, the space allows users to set custom system messages for each assistant, potentially completely changing the workflow. # Workflow Steps 1. User Input: - The user defines a prompt, such as "Write a random password generator using python." - Outcome: A clear task for the primary assistant to accomplish. 2. Primary Assistant Work: - The primary assistant begins working on the provided prompt. It generates an initial code snippet based on the user's request. - Outcome: An initial proposal for the requested code. 3. Critic Feedback: - The critic reviews the generated code provides feedback or (if the output meets the criteria), broadcasts the APPROVED message. (This process repeats until the output is APPROVED or 10 messages have been exchanged). - Outcome: A revised Python function that incorporates the critic's feedback. 4. Documentation Generation: - Once the code is approved, it is passed to a documentation assistant. The documentation assistant generates a concise documentation for the final code. - Outcome: A short documentation including function description, parameters, and return values. Enjoy! https://huggingface.co/spaces/CultriX/AutoGen-MultiAgent-Example

reacted to CultriX's post with ❤️ about 22 hours ago

# Space for Multi-Agent Workflows using AutoGen Hi all, I created this "AutoGen Multi-Agent Workflow" space that allows you to experiment with multi-agent workflows. By default, it allows code generation with built-in quality control and automatic documentation generation. It achieves this by leveraging multiple AI agents working together to produce high-quality code snippets, ensuring they meet the specified requirements. In addition to the default, the space allows users to set custom system messages for each assistant, potentially completely changing the workflow. # Workflow Steps 1. User Input: - The user defines a prompt, such as "Write a random password generator using python." - Outcome: A clear task for the primary assistant to accomplish. 2. Primary Assistant Work: - The primary assistant begins working on the provided prompt. It generates an initial code snippet based on the user's request. - Outcome: An initial proposal for the requested code. 3. Critic Feedback: - The critic reviews the generated code provides feedback or (if the output meets the criteria), broadcasts the APPROVED message. (This process repeats until the output is APPROVED or 10 messages have been exchanged). - Outcome: A revised Python function that incorporates the critic's feedback. 4. Documentation Generation: - Once the code is approved, it is passed to a documentation assistant. The documentation assistant generates a concise documentation for the final code. - Outcome: A short documentation including function description, parameters, and return values. Enjoy! https://huggingface.co/spaces/CultriX/AutoGen-MultiAgent-Example

View all activity

Articles

Decoding Strategies in Large Language Models

Fine-tune Llama 3.1 Ultra-Efficiently with Unsloth

The Rise of Agentic Data Generation

Uncensor any LLM with abliteration

Fine-tune Llama 3 with ORPO

Create Mixtures of Experts with MergeKit

Merge Large Language Models with mergekit

Organizations

mlabonne's activity

upvoted a paper about 18 hours ago

Enhancing Human-Like Responses in Large Language Models

Paper • 2501.05032 • Published 2 days ago • 11

upvoted 2 articles about 2 months ago

Article

The Beginners Guide to Cleaning a Dataset

By

•

Nov 18, 2024

• 24

Article

Releasing the largest multilingual open pretraining dataset

By

•

Nov 13, 2024

• 98

upvoted an article 2 months ago

Article

Decoding Strategies in Large Language Models

By

•

Oct 29, 2024

• 38

upvoted a paper 3 months ago

Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs

Paper • 2402.14740 • Published Feb 22, 2024 • 12

upvoted an article 3 months ago

Article

VLM Art Analysis

By

•

Oct 4, 2024

• 11

upvoted an article 4 months ago

Article

Fine-tuning LLMs to 1.58bit: extreme quantization made easy

Sep 18, 2024

• 215

upvoted a collection 5 months ago

🧠 Abliteration

Uncensored models using abliteration. See this article for more information: huggingface.co/blog/mlabonne/abliteration • 7 items • Updated Nov 18, 2024 • 24

upvoted an article 5 months ago

Article

Introduction to ggml

Aug 13, 2024

• 125

upvoted a paper 5 months ago

The Impact of Hyperparameters on Large Language Model Inference Performance: An Evaluation of vLLM and HuggingFace Pipelines

Paper • 2408.01050 • Published Aug 2, 2024 • 8

upvoted an article 5 months ago

Article

The case for specialized pre-training: ultra-fast foundation models for dedicated tasks

By

•

Aug 4, 2024

• 27

upvoted a paper 5 months ago

Improving Text Embeddings for Smaller Language Models Using Contrastive Fine-tuning

Paper • 2408.00690 • Published Aug 1, 2024 • 24

upvoted a collection 5 months ago

Probably function calling datasets

Created using the https://huggingface.co/spaces/librarian-bots/dataset-column-search-api Space. • 39 items • Updated Jul 17, 2024 • 36

upvoted 2 papers 6 months ago

Executable Code Actions Elicit Better LLM Agents

Paper • 2402.01030 • Published Feb 1, 2024 • 40

Understanding Reference Policies in Direct Preference Optimization

Paper • 2407.13709 • Published Jul 18, 2024 • 16

upvoted 3 collections 6 months ago

Bad Data Toolbox

PleIAs collection of models for the data processing of challenging document and data sources. • 5 items • Updated Jul 18, 2024 • 15

OpenCulture

A multilingual dataset of public domain books and newspapers. • 27 items • Updated Nov 6, 2024 • 122

Finance Commons

A large collection of multimodal financial documents in open data. • 7 items • Updated Jul 17, 2024 • 7

upvoted an article 6 months ago

Article

Fine-tune Llama 3.1 Ultra-Efficiently with Unsloth

By

•

Jul 29, 2024

• 262

upvoted a paper 6 months ago

The Importance of Online Data: Understanding Preference Fine-tuning via Coverage

Paper • 2406.01462 • Published Jun 3, 2024 • 6