Add new Relax function to the batched model for evaluating query tokens over multiple time steps in parallel #156

masahi merged 7 commits intooctoml:batch-servingoctoml/mlc-llm:batch-servingfrom

masahi:model-multi-query-logitsmasahi/mlc-llm:model-multi-query-logitsCopy head branch name to clipboard