eliebak (Elie Bakouch)

upvoted a paper 8 days ago

Gemma 3 Technical Report

Paper • 2503.19786 • Published 10 days ago • 41

upvoted a paper 11 days ago

Reinforcement Learning for Reasoning in Small LLMs: What Works and What Doesn't

Paper • 2503.16219 • Published 15 days ago • 46

upvoted an article 17 days ago

Article

Xet is on the Hub

18 days ago

• 41

upvoted a paper about 1 month ago

SOAP: Improving and Stabilizing Shampoo using Adam

Paper • 2409.11321 • Published Sep 17, 2024 • 1

upvoted a collection about 1 month ago

🧩 SmolLM2 Intermediate Checkpoints

Collection

3 items • Updated 5 days ago • 3

upvoted a paper about 1 month ago

Small Models Struggle to Learn from Strong Reasoners

Paper • 2502.12143 • Published Feb 17 • 33

upvoted a collection about 1 month ago

PLLuM-chat

Collection

6 items • Updated Feb 24 • 8

upvoted an article about 1 month ago

Article

SmolVLM2: Bringing Video Understanding to Every Device

Feb 20

• 223

upvoted a collection about 1 month ago

Granite Data

Collection

This collection has a set of artifacts which are related to curating and evaluating datasets used for Granite models • 16 items • Updated Feb 28 • 4

upvoted an article about 2 months ago

Article

Introducing Three New Serverless Inference Providers: Hyperbolic, Nebius AI Studio, and Novita 🔥

Feb 18

• 95

upvoted 4 papers about 2 months ago

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Paper • 2501.18965 • Published Jan 31 • 7

upvoted an article about 2 months ago

Article

Open-source DeepResearch – Freeing our search agents

Feb 4

• 1.21k

Elie Bakouch

AI & ML interests

Organizations

eliebak's activity

Gemma 3 Technical Report

Reinforcement Learning for Reasoning in Small LLMs: What Works and What Doesn't

Xet is on the Hub

SOAP: Improving and Stabilizing Shampoo using Adam

🧩 SmolLM2 Intermediate Checkpoints

Small Models Struggle to Learn from Strong Reasoners

PLLuM-chat

SmolVLM2: Bringing Video Understanding to Every Device

Granite Data

Introducing Three New Serverless Inference Providers: Hyperbolic, Nebius AI Studio, and Novita 🔥

INTELLECT-1 Technical Report

On Teacher Hacking in Language Model Distillation

SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Open-source DeepResearch – Freeing our search agents