Predictable Scaling and Infrastructure
•What ∘Collect a dataset of text, code, images, audio, math. ∘Pre-train a model to predict the next word. ∘Post-train it to e.g. follow instructions, be conversational, use tools. •How ∘Synchronous SGD of a transformer on a large cluster of accelerators. ∘Combine many forms of data and model parallelism. ∘Reinforcement learning from human feedback.
Authors
- Trevor Cai
Publication Details
- Published
- 2024-08-25
- DOI
- https://doi.org/10.1109/hcs61935.2024.10664929
- Citations
- 1
- Primary Topic
- Generative Adversarial Networks and Image Synthesis
- Type
- article
- Field-Weighted Citation Impact
- 0.25