Adapter-Augmented Bandits for Online Multi-Constrained Multi-Modal Inference Scheduling
This work addresses scheduling challenges for MLLM inference in practical systems with varying costs and constraints, offering an incremental improvement over existing methods.
The paper tackles the problem of online multi-modal large language model (MLLM) inference scheduling under heterogeneous budgets and uncertainties, proposing M-CMAB, which achieves up to 14.18% higher reward than state-of-the-art baselines.
Multi-modal large language model (MLLM) inference scheduling enables strong response quality under practical and heterogeneous budgets, beyond what a homogeneous single-backend setting can offer. Yet online MLLM task scheduling is nontrivial, as requests vary sharply in modality composition and latent reasoning difficulty, while execution backends incur distinct, time-varying costs due to system jitter and network variation. These coupled uncertainties pose two core challenges: deriving semantically faithful yet scheduling-relevant multi-modal task representations, and making low-overhead online decisions over irreversible multi-dimensional budgets. Accordingly, we propose \emph{M-CMAB} (\underline{M}ulti-modal \underline{M}ulti-constraint \underline{C}ontextual \underline{M}ulti-\underline{A}rmed \underline{B}andit), a multi-adapter-enhanced MLLM inference scheduling framework with three components: (i) a CLS-attentive, frozen-backbone \emph{Predictor} that extracts compact task representations and updates only lightweight adapters for action-specific estimation; (ii) a primal-dual \emph{Constrainer} that maintains online Lagrange multipliers to enforce long-horizon constraints via per-round objectives; and (iii) a two-phase \emph{Scheduler} that balances exploration and exploitation under irreversible budgets. We establish a regret guarantee under multi-dimensional knapsack constraints. On a composite multimodal benchmark with heterogeneous backends, \emph{M-CMAB} consistently outperforms state-of-the-art baselines across budget regimes, achieving up to 14.18% higher reward and closely tracking an oracle-aided upper bound. Codes are available at https://anonymous.4open.science/r/M2CMAB/.