ES-FoMo@ICML2025
@ESFoMo
Followers
356
Following
52
Media
29
Statuses
55
Efficient Systems for Foundation Models Workshop, ICML2025. Join us if you are interested in the challenges associated with large models training & inference!
Vancouver, Canada
Joined March 2023
@MatharyCharles @_albertgu And presenting a Best Poster award to "Unified Scaling Laws for Compressed Representations" by Andrei Panferov, Alexandra Volkova, @ionutmodo, Vage Egiazarian, @mher_safaryan, and @DAlistarh! Congrats!
0
3
10
And now a talk on hardware-efficient attention for decoding!
1
0
0
0
0
3
1
0
2
1
0
2
1
0
0
@wanchao_ @bariskasikci Our first contributed talk of the day - FPTQuant for outlier-aware quantization!
1
0
1
@wanchao_ Next we have @bariskasikci with a talk on the quest for blazingly fast LLM inference!
1
0
0
Our first talk of the day - @hagay_lupesko on scaling from zero to 50 exaflops in a year!
Looking forward to seeing everyone for ES-FoMo part three tomorrow! We'll be in East Exhibition Hall A (the big one), and we've got an exciting schedule of invited talks, orals, and posters planned for you tomorrow. Let's meet some of our great speakers! 1/
3
1
12
@realDanFu @slippylolo @m_ryabinin @DanielHesslow @simran_s_arora @SonglinYang4 @dan_biderman @tri_dao @BeidiChen @Azaliamirh @percyliang And… just one more thing! @_albertgu will be joining us at 4:30pm to give a talk about his new work on H-nets. Come hear about his takedown of big token! 12/11
1
0
14
Finally, thanks to all the organizers who made this event possible - @realDanFu, @slippylolo, @m_ryabinin, @DanielHesslow, @simran_s_arora, @SonglinYang4, @dan_biderman, @tri_dao, @BeidiChen, @Azaliamirh, and @percyliang! We're looking forward to welcoming you tomorrow! 11/11
1
0
7
And last but not least, we'll end the day with a talk from @Google's @MatharyCharles, who will be talking to us about Making Communication-Efficient Language Model Training Reliable and Scalable with DiLoCo. We're looking forward to it! 10/
1
3
2
After a short coffee break, we'll have our second oral session of the day, at 3:30pm. This one will feature two great papers: Any-Order GPT as Masked Diffusion Model, and Hardware-Efficient Attention for Fast Decoding. Check them out! 9/
1
0
2
Then, @Avanika15 (@StanfordAILab / @rox__ai) will bring us back to our regularly scheduled programming with a talk on Minions: Cost-efficient Collaboration Between On-device and Cloud Language Models, at 2:30pm. Local + cloud, exciting! 8/
1
1
5
After a quick lunch break (or perhaps a last-minute talk by the elusive @_albertgu... stay tuned!), we'll move on to the highlight of the day - the poster session from 1:00-2:30pm! This year's session is bigger than ever - we'll have 150 papers! You won't want to miss it! 7/
1
0
6