Maxime Labonne PRO

mlabonne

517 181 1245

https://mlabonne.github.io/blog

AI & ML interests

Post-training, model editing, quantization

Recent Activity

updated a model 3 days ago

LiquidAI/LFM2.5-2.6B-Base

updated a collection 5 days ago

👁️ LFM2.5-VL

updated a collection 5 days ago

👁️ LFM2.5-VL

View all activity

Organizations

upvoted an article 5 days ago

Article

LFM2.5-VL-3B for Better and Faster Vision Capabilities for the Edge

LiquidAI

•

5 days ago

• 40

upvoted an article 12 days ago

Article

Deploy local agents everywhere with LFM2.5-2.6B

LiquidAI

•

13 days ago

• 90

upvoted an article 18 days ago

Article

mDenseOn with the mLateOn: Open Multilingual, Long-Context, and Code Retrieval Models

lightonai

•

18 days ago

• 36

upvoted an article 20 days ago

Article

LFM2.5-Encoders for Fast Long-Context Inference on CPU

LiquidAI

•

20 days ago

• 68

upvoted a collection about 2 months ago

BidirLM

Collection

BidirLM is a family of 5 frontier bidirectional encoders, including an omnimodal variant at 2.5B. • 8 items • Updated Apr 15 • 4

upvoted an article 4 months ago

Article

Keep the Tokens Flowing: Lessons from 16 Open-Source RL Libraries

aminediroHF, qgallouedec, kashif, lewtun, edbeeching, albertvillanova, nouamanetazi, lvwerra, sergiopaniego

•

Mar 10

• 175

upvoted 7 articles 6 months ago

Article

Kimi K2.5: Still Worth It After Two Weeks?

mlabonne

•

Feb 23

• 10

Article

GGML and llama.cpp join HF to ensure the long-term progress of Local AI

ggerganov, ngxson, allozaur, lysandre, victor, julien-c

•

Feb 20

• 509

Article

Train AI models with Unsloth and Hugging Face Jobs for FREE

burtenshaw, danielhanchen, shimmyshimmer, mlabonne, davanstrien, evalstate

•

Feb 20

• 107

Article

Qwen3.5: Nobody Agrees on Attention Anymore

mlabonne

•

Feb 17

• 27

Article

MiniMax-M2.5: The $1/hour Frontier Model

mlabonne

•

Feb 17

• 1

Article

GLM-5: China's First Public AI Company Ships a Frontier Model

mlabonne

•

Feb 17

• 3

Article

Community Evals: Because we're done trusting black-box leaderboards over the community

burtenshaw, SaylorTwift, kramp, merve, davanstrien, nielsr, julien-c

•

Feb 4

• 90

upvoted 3 articles 7 months ago

Article

Why Your AI Strategy Needs Hugging Face Storage

AdrianLepers

•

Jan 26

• 13

Article

Open Responses: What you need to know

evalstate, burtenshaw, merve, pcuenq

•

Jan 15

• 112

Article

SmolLM-Smashed: Tiny Giants, Optimized for Speed

PrunaAI

•

Jan 13

• 15

upvoted 4 collections 7 months ago

Maxime Labonne PRO

AI & ML interests

Recent Activity

Organizations

mlabonne's activity

LFM2.5-VL-3B for Better and Faster Vision Capabilities for the Edge

Deploy local agents everywhere with LFM2.5-2.6B

mDenseOn with the mLateOn: Open Multilingual, Long-Context, and Code Retrieval Models

LFM2.5-Encoders for Fast Long-Context Inference on CPU

Keep the Tokens Flowing: Lessons from 16 Open-Source RL Libraries

Kimi K2.5: Still Worth It After Two Weeks?

GGML and llama.cpp join HF to ensure the long-term progress of Local AI

Train AI models with Unsloth and Hugging Face Jobs for FREE

Qwen3.5: Nobody Agrees on Attention Anymore

MiniMax-M2.5: The $1/hour Frontier Model

GLM-5: China's First Public AI Company Ships a Frontier Model

Community Evals: Because we're done trusting black-box leaderboards over the community

Why Your AI Strategy Needs Hugging Face Storage

Open Responses: What you need to know

SmolLM-Smashed: Tiny Giants, Optimized for Speed