diff --git a/python/tabby-loadtest/README.md b/python/tabby-loadtest/README.md new file mode 100644 index 000000000000..1fd64d1cc987 --- /dev/null +++ b/python/tabby-loadtest/README.md @@ -0,0 +1,6 @@ +# Run loadtest with tabby on modal GPUs + +Steps: +1. Adjust `TABBY_API_HOST` in `run.sh` to match your modal deployment url. +2. Add models you're interested in to benchmark at end of `run.sh` +3. Run `run.sh`, output will be appended to `record.csv` \ No newline at end of file