Lewis Tunstall PRO

lewtun's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes

upvoted a paper 4 days ago

A Simple and Provable Scaling Law for the Test-Time Compute of Large Language Models

Paper • 2411.19477 • Published Nov 29, 2024 • 6

liked a Space 5 days ago

Running

30

🥇

MEGA-Bench

A leaderboard for multimodal models

liked a model 5 days ago

HuggingFaceTB/FineMath-Llama-3B

Updated 5 days ago • 126 • 12

reacted to prithivMLmods's post with 🚀 5 days ago

view post

Post

5372

Reasoning SmolLM2 🚀

🎯Fine-tuning SmolLM2 on a lightweight synthetic reasoning dataset for reasoning-specific tasks. Future updates will focus on lightweight, blazing-fast reasoning models. Until then, check out the blog for fine-tuning details.

🔥Blog : https://huggingface.co./blog/prithivMLmods/smollm2-ft

🔼 Models :
+ SmolLM2-CoT-360M : prithivMLmods/SmolLM2-CoT-360M
+ Reasoning-SmolLM2-135M : prithivMLmods/Reasoning-SmolLM2-135M
+ SmolLM2-CoT-360M-GGUF : prithivMLmods/SmolLM2-CoT-360M-GGUF

🤠 Other Details :
+ Demo : prithivMLmods/SmolLM2-CoT-360M
+ Fine-tune nB : prithivMLmods/SmolLM2-CoT-360M

See translation

• 

🚀

18

18

🔥

11

11

+

Reply

updated a collection 6 days ago

Scaling Test-Time Compute with Open Models

Collection

Models and datasets used in our blog post: https://huggingface.co./spaces/HuggingFaceH4/blogpost-scaling-test-time-compute • 10 items • Updated 6 days ago • 21

updated a dataset 6 days ago

lewtun/Llama-3.2-1B-Instruct-best_of_n-prm-completions

Viewer • Updated 6 days ago • 10 • 12

liked a dataset 6 days ago

HuggingFaceH4/MATH-500

Viewer • Updated Nov 15, 2024 • 500 • 7.1k • 39

liked a model 6 days ago

deepseek-ai/DeepSeek-V3

Updated 13 days ago • 110k • 1.66k

posted an update 6 days ago

view post

Post

3162

I was initially pretty sceptical about Meta's Coconut paper [1] because the largest perf gains were reported on toy linguistic problems. However, these results on machine translation are pretty impressive!

https://x.com/casper_hansen_/status/1875872309996855343

Together with the recent PRIME method [2] for scaling RL, reasoning for open models is looking pretty exciting for 2025!

[1] Training Large Language Models to Reason in a Continuous Latent Space (2412.06769)
[2] https://huggingface.co./blog/ganqu/prime

See translation

• 

🔥

8

8

🧠

2

2

+

Reply

updated a collection 6 days ago

Scaling Test-Time Compute with Open Models

Collection

Models and datasets used in our blog post: https://huggingface.co./spaces/HuggingFaceH4/blogpost-scaling-test-time-compute • 10 items • Updated 6 days ago • 21

updated a Space 9 days ago

Running

457

📈

Scaling test-time compute

New activity in HuggingFaceH4/blogpost-scaling-test-time-compute 9 days ago

Questions about Verifier Development, Search as Data Generation Tool, and Model Family Alignment

4

#8 opened 22 days ago by

bird-of-paradise

liked a model 9 days ago

Skywork/Skywork-o1-Open-PRM-Qwen-2.5-1.5B

Text Classification • Updated Nov 27, 2024 • 926 • 22

New activity in HuggingFaceH4/blogpost-scaling-test-time-compute 10 days ago

Link to "canonical form" does not work

1

#4 opened 26 days ago by

blackhc

code pointers?

2

#7 opened 22 days ago by

rganti

Is there a way to print this article?

1

#9 opened 18 days ago by

iamgianluca