Scalable Distributed Stochastic Optimization via Bidirectional Compression: Beyond Pessimistic Limits
In centralized, distributed, and federated learning with stochastic gradients and $n$ workers, it was recently shown that it is infeasible to find an $\varepsilon$-stationary point faster than $\tildeΩ\left(\min\left\{\frac{d κL Δ}{\varepsilon} + \frac{h L Δ}{\varepsilon} + \frac{h σ^2 L Δ}{n \varepsilon^2},\; \frac{h σ^2 L Δ}{\varepsilon^2} + \frac{h L Δ}{\varepsilon}\right\}\right)$ seconds in both homogeneous and heterogeneous settings under standard assumptions: $L$-smoothness, $σ^2$-bounded unbiased stochastic gradients, and lower boundedness of the function, i.e., $f(x) \geq f^*$ for…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2605.07795
curl -X POST /api/v1/records/arxiv/2605.07795/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2605.07795/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'