Batch-Max: Higher LLM Throughput using Larger Batch Sizes and KV Cache Compression

Add code
Dec 07, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: