OPML: Optimistic Machine Learning on Blockchain - Layer 2 - Ethereum Research
TL;DR We propose OPML (Optimistic Machine Learning), which enables AI model inference and training/fine-tuning on the blockchain system using optimistic approach (OPML is also called FPML, where “FP” refers to fraud proof). OPML can provide ML service with low cost and high efficiency compared to ZKML. The participation requirement for OPML is low: We are now able to run OPML with a large language model, e.g., 7B-LLaMA (the model size is around 26GB) on a common PC without GPU. OPML adopts a v...
OPML: Optimistic Machine Learning on Blockchain - Layer 2 - Ethereum Research OPML: Optimistic Machine Learning on Blockchain Layer 2 0x1cc July 31, 2023, 2:29pm 1 TL;DR We propose OPML (Optimistic Machine Learning), which enables AI model inference and training/fine-tuning on the blockchain system using optimistic approach (OPML is also called FPML , where “FP” refers to fraud proof). OPML can provide ML service with low cost and high efficiency compared to ZKML. The participation requirement for OPML is low: We are now able to run OPML with a large language model, e.g., 7B-LLaMA (the model s
Explore this link on the map →related reading
- Pearl Whitepaperpearlresearch.ai
- ZKML Research Initiatives - HackMDhackmd.io
- Scaling up ZK for DNN Inferencearxiv.org
- LLM Inference Performance Engineering: Best Practices | Databricks Blogdatabricks.com
- Where Zero-Knowledge Machine Learning (zkML) Fits into the Bigger Picture of AIinsidejuice.substack.com
- ZK Machine Learning - HackMDhackmd.io
- Machine learning and zero-knowledge proofs - a16z cryptoa16zcrypto.com
- An introduction to zero-knowledge machine learning (ZKML)worldcoin.org
- Reiner Pope – The math behind how LLMs are trained and serveddwarkesh.com
- Sumcheck good. Lookups good. JOLT good. Particularly for zero-knowledge machine learning.blog.icme.io
- o1 and Reasoning | AndoLogsblog.ando.ai
- Estimating training compute of deep learning models | Epoch AIepochai.org