Day 1:  Thu, March 28th 

morning session:   Agora:  9:30am - 1:00pm

9:45  - 10:45 An Intro:  On Scaling  Laws, Foundation Models and HPC   Irina Rish (CERC-AAI Lab Lead, UdeM/Mila)    slides / video

 An overview of Large-Scale AI Projects at CERC-AAI Lab and AGI Collective 

break/coffee/Q&A

11:00 - 12:00 A tutorial on HPC & Building Foundation Models Quentin Anthony (CERC-AAI/OSU/Eleuther/Zyphra) slides / video

For more background, you can also see tutorials at our course  webpage  HPC, Large-Scale Models 

12:00 - 1:00 Cerebras Systems: Hardware, Software, Scaling Laws and Sparsity (Natalia Vassilieva, Joel Hestness)   slides / video  (starting at: 01:01:53)

lunch

afternoon session:   A14, Mila 6666:  2:00pm - 5:00pm

2:00 -5:00 A tutorial on HPC & Building Foundation Models:  Part 2  Quentin Anthony slides / video 1, video 2, video  3

__________________

Day 2: Fri, March 29th  (Agora)

In-depth Overview of Large-Scale AI Projects at CERC-AAI Lab and AGI Collective; Discussion of Open Questions, Next Steps, and Collaborations

10:00 - 10:30 Time-Series Foundation Models Arjun Ashok, Roland Riachi    slides / video  (paper, blog, tweet) 

10:30- 11:00 Continual Pretraining of Foundation Models   Benjamin ThérienAdam IbrahimKshitij Gupta slides / video (paper,  blog,  tweet)   

  

break/coffee/Q&A

11:10 - 11:40  Robin Suite of Open-Source Multimodal Foundation Models   Kshitij Gupta,  Daniel Kaplan   slides / video (paper, blog, tweet)

11:40 - 12:10  Multimodal Alignment: Towards ethical multimodal systems    Alexis Roger   slides  / video (starting at 00:18:58(paper1, paper2)

12:10 - 12:30 The New Frontier of Generated Data Diganta Misra slides / video  (paper, tweet)

lunch (pizza)

1:00 - 1:30 LLM 4 Psychology and Psychology 4 LLMs Tommaso Tosato 

1:30 - 2:00  Compression and Fast Inference in Foundation Models   Tejas Vaidhya slides / video (paper , blog, tweet)

break/coffee/Q&A

2:00 - 3:00 LLMs through the HPC Lens Quentin Anthony (CERC-AAAI Lab/OSU/Eleuther/Zyphra) slides / video

3:00 - 5:00 Q&A with Quentin: in person and online

Relevant materilas from previoius scaling workshop:
an overview of large-scale AI projects at CERC-AAI Lab and AGI Collective presented in Dec at the  6th Neural Scaling Workshop  in New Orleans, co-located with NeurIPS 2023 

Ayush Kaushal Nolano: Compression and Fast Inference in Foundation Models  (8 min)   (paper , blog, tweet, slides)

Ayush Kaushal:  Nolano: Introducing  Hi-NOLIN - the First Hindi-English LLM   (5 min) ( tweet, blog, slides)