Sylvain Gugger
@GuggerSylvain
Followers
27K
Following
2K
Media
98
Statuses
1K
Machine Learning at Jane Street. Previously at @huggingface and @fastdotai Co-author of https://t.co/lywnOAwwnc He/him
Brooklyn, NY
Joined March 2018
One year and half after starting the first draft of the first chapter, look what arrived in the mail!
48
120
1K
Very excited to collaborate with Mark on this!.
On Sep 6 in NYC, this won't be your typical hackathon where you do your own thing in a corner and then present at the of the day. You'll deploy real models to the market, trades will happen, chaos should be expected. The fastest model is great but time to market matters more.
1
2
16
RT @LysandreJik: The new transformers release comes w/ a surprise: kernels support ⚡️. It integrates deeply with precompiled kernels on the….
0
17
0
RT @WentaoGuo7: 🦆🚀QuACK🦆🚀: new SOL mem-bound kernel library without a single line of CUDA C++ all straight in Python thanks to CuTe-DSL. On….
0
73
0
I’ll be giving a talk on basic GPU performance in SF next week. Come and say hi if you’d like to learn more about the kind of machine learning we do at Jane Street!
9
17
386
RT @Thom_Wolf: Thrilled to finally share what we've been working on for months at @huggingface 🤝@pollenrobotics. Our first robot: Reachy Mi….
0
516
0
RT @__tensorcore__: 🚨🔥 CUTLASS 4.0 is released 🔥🚨. pip install nvidia-cutlass-dsl. 4.0 marks a major shift for CUTLASS: towards native GPU….
0
86
0
RT @joao_gante: Speculative Decoding before: limited choices, the draft model must have the same tokenizer 😬.Speculative Decoding now: unli….
0
9
0
There is also a new CUTLASS Python library that matches the C++ performance!
2
5
44
I’m especially excited about cuTile which looks like a new easy way to program tensor cores without sacrificing performance.
3
5
39
Looks like 2025 is going to be the year of CUDA in Python, new libraries at all level of the stack!
2
25
226
RT @bfspector: (1/7) Inspired by DeepSeek's FlashMLA, we're releasing ThunderMLA—a fused megakernel optimized for variable-prompt decoding!….
0
71
0
RT @GPU_MODE: Write a fast kernel and run it on Discord. See how you compare against the best!. If you're familiar with Leetcode, Kaggle or….
0
40
0
RT @Nouamanetazi: 🚀 Excited to release *THE* Ultra-Scale Playbook - a comprehensive guide on training LLMs from 1 to 1000s of GPUs! https:/….
0
233
0
RT @StasBekman: This is huge, huge, huge - DeepSpeed is now a community-owned project as it's now a part of the Linux Foundation. Committer….
0
17
0
RT @morgymcg: TIL Jane Street have an eng podcast. Most recent episode is with @GuggerSylvain on training & ML infra.
0
2
0
We had an awesome talk at Jane Street from the amazing @cHHillee on scaling ML systems to and I just realized the recording is now online:
4
44
453
RT @cHHillee: Jane Street tech talks have always been super awesome. So I'm quite excited to be visiting Jane Street on Monday to give a ta….
0
31
0