Steven @PlsGoForIt

Code Don't be a fool Stockholm, Sverige Joined February 2023

Tweets

1K
Followers

110
Following

1K
Likes

5K

九原客 @9hills

2 days ago

免费 GPU 或廉价算力推荐，可用于学习10B以内模型微调训练，我会给学员推荐这些环境练习模型训练。 github.com/ninehills/blog…

3 91 421 35K 388

Download Image

Krishna Mohan @KMohan2006

4 days ago

Good morning with a banger resource on vLLM

4 42 521 122K 604

Download Image

Rohit Kumar Tiwari @_rohit_tiwari_

2 weeks ago

The Ultimate Guide to Fine-Tuning LLMs. Here’s a 7-step roadmap that makes it simple. This technical report takes a deep look at the fine-tuning process for LLMs. Combining both theory and practice. 1. Introduction 2. Seven Stage Fine-Tuning Pipeline ↳ Stage-1: Data…

133 113 604 83K 973

Download Image

You Jiacheng @YouJiacheng

2 weeks ago

btw it's possible to use mantissa-free weights. The Chief Scientist of NVIDIA had a keynote with content about it. nvidia.com/en-us/on-deman…

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTex

2 weeks ago

btw it's possible to use mantissa-free weights. The Chief Scientist of NVIDIA had a keynote with content about it. nvidia.com/en-us/on-deman… https://t.co/5wIfpJ6Vk8

4 4 74 24K 31

Download Image

3 15 117 11K 79

Download Image

Tanishq Mathew Abraham, Ph.D. @iScienceLuvr

2 weeks ago

RL Is Neither a Panacea Nor a Mirage: Understanding Supervised vs. Reinforcement Learning Fine-Tuning for LLMs "RL primarily counteracts SFT-induced directional drift rather than finding new solutions. Our spectrum-aware analysis highlights inexpensive recovery knobs low-rank…

6 57 311 22K 294

Download Image

SemiAnalysis @SemiAnalysis_

2 weeks ago

TogetherAI's Chief Scientist @tri_dao announced Flash Attention v4 at HotChips Conference which is up to 22% faster than the attention kernel implementation from NVIDIA's cuDNN library. Tri Dao was able to achieve this 2 key algorithmic changes. Firstly, it uses a new online…

24 76 627 98K 225

Download Image

Mohit Mishra @chessMan786

3 weeks ago

Understanding the Linux Virtual Memory Manager

3 59 513 16K 356

Download Image

Guangxuan Xiao @Guangxuan_Xiao

4 weeks ago

I've written the full story of Attention Sinks — a technical deep-dive into how the mechanism was developed and how our research ended up being used in OpenAI's new OSS models. For those interested in the details: hanlab.mit.edu/blog/streaming…

39 275 2K 247K 2K

Download Image

himanshu @himanshustwts

4 weeks ago

just published a curated list of amazing blogs/articles on ai. highly selective. feel free to comment if you find anything interesting, will update it async. (link in replies)

12 43 529 35K 850

Download Image

AK @_akhaliq

4 weeks ago

R-Zero Self-Evolving Reasoning LLM from Zero Data

13 84 554 63K 487

Download Image

Aran Komatsuzaki @arankomatsuzaki

2 months ago

Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety

7 11 69 36K 45

Download Image

Underfox @Underfox3

2 months ago

In this paper is presented a detailed experimental analysis of NVIDIA’s Blackwell architecture through microbenchmarks with a comparison to the previous Hopper generation GPUs. arxiv.org/pdf/2507.10789

4 26 125 7K 78

Download Image

ℏεsam @Hesamation

3 months ago

This is a solid 29 videos playlist on how to build DeepSeek from scratch. It covers theory and code, from the very foundations to advanced. Self attention, multi-head [latent] attention, GQA, how DeepSeek rewrote Quantization, etc. One video a day and you’ll finish in a month.

6 203 1K 75K 2K

Download Image

Aran Komatsuzaki @arankomatsuzaki

3 months ago

Gemini 2.5 tech report was released!

15 89 512 55K 246

Download Image

typedfemale @typedfemale

3 months ago

a paper describing triton's linear layouts is out!

5 29 364 63K 246

Download Image

Damek @damekdavis

3 months ago

i wrote a short post on this. i really shouldn't have, but I learned a lot. i wanted to cover more like "optimal swizzling", but i ran out of time.

typedfemale @typedfemale

3 months ago

i wrote a short post on this. i really shouldn't have, but I learned a lot. i wanted to cover more like "optimal swizzling", but i ran out of time. https://t.co/rHFKJc6WJX

5 29 364 63K 246

Download Image

4 32 336 32K 377

Download Image

Vijay @tensorcore

3 months ago

Another 🔥 blog about CUTLASS from @colfaxintl, this time focusing on the gory details of block-scaled MXFP and NVFP data types and Blackwell kernels for them. research.colfax-intl.com/cutlass-tutori…

0 36 158 12K 81

Dan | Machine Learning Engineer @DanKornas

3 months ago

Learn the Basic Fundamentals of Pytorch

1 187 1K 62K 1K

Download Image

Vivek Galatage @vivekgalatage

3 months ago

If you are beginning with compilers and are interested in clang, "An Overview of Clang" is a good talk. youtu.be/5kkMpJpIGYU

2 55 427 14K 428

Download Image

马东锡 NLP @dongxi_nlp

3 months ago

「RLVR, Table Reasoning」 Table-R1: Inference-Time Scaling for Table Reasoning Inference-Time Scaling 并非数学或coding的专属，在高结构化数据推理中，也可把“想得更久”转化为性能。跳出math和coding的盒子，作者首次将 ITS 系统性地迁移到结构化数据（如 tabular data）推理的研究。…