A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 22:00 UTC
Feature/add benchmark testing (#5947)
Co-authored-by: Roger Wang <ywang@roblox.com>
Haichuan committed
Jul 8, 2024 at 15:52 UTC
717f4bcea036a049e86802b3a05dd6f7cd17efc8
1 file changed
+3
-3
benchmarks/benchmark_serving.py
+3
-3
@@ -483,10 +483,10 @@ def main(args: argparse.Namespace):
483
484
elif args.dataset_name == "random":
485
input_requests = sample_random_requests(
486
- input_len=args.input_len,
487
- output_len=args.output_len,
486
+ input_len=args.random_input_len,
487
+ output_len=args.random_output_len,
488
num_prompts=args.num_prompts,
489
- range_ratio=args.range_ratio,
489
+ range_ratio=args.random_range_ratio,
490
tokenizer=tokenizer,
491
)
492