context-length

Here is 1 public repository matching this topic...

chenghanshih / ctxprobe

Find the largest context a GGUF model actually runs at on one GPU. Boots llama-server for real, climbs a 64/512/4096-token prompt ladder, validates the winner at 95% fill. Measured ceilings and the gotchas behind them, EN/繁中.

benchmark cuda vram kv-cache llama-cpp local-llm llm-inference context-length gguf

Updated Sep 11, 2026
Shell

Add this topic to your repo

To associate your repository with the context-length topic, visit your repo's landing page and select "manage topics."

Learn more