963fcfa48c
Co-authored-by: Claude Fable 5 <noreply@anthropic.com> Co-authored-by: Bugen Zhao <i@bugenzhao.com> Signed-off-by: esmeetu <jasonailu87@gmail.com> Signed-off-by: Bugen Zhao <i@bugenzhao.com>
111 lines
2.3 KiB
Plaintext
111 lines
2.3 KiB
Plaintext
# Long CLI flags of the Python `vllm bench serve` parser
|
|
# (vllm.benchmarks.serve.add_cli_args), sorted; --help excluded.
|
|
# Regenerate: .venv/bin/python tests/benchmarks/test_rust_bench_cli_parity.py
|
|
# Consumed by rust/src/bench/tests/cli_parity.rs, which requires every flag
|
|
# below to be a known vllm-bench flag/alias or a PYTHON_ONLY allowlist entry.
|
|
--append-result
|
|
--asr-max-audio-len-sec
|
|
--asr-min-audio-len-sec
|
|
--backend
|
|
--base-url
|
|
--bfcl-categories
|
|
--blazedit-max-distance
|
|
--blazedit-min-distance
|
|
--burstiness
|
|
--chat-template-kwargs
|
|
--custom-ensure-client-side-data
|
|
--custom-output-len
|
|
--dataset-name
|
|
--dataset-path
|
|
--disable-shuffle
|
|
--disable-tqdm
|
|
--enable-multimodal-chat
|
|
--endpoint
|
|
--extra-body
|
|
--frequency-penalty
|
|
--goodput
|
|
--header
|
|
--hf-name
|
|
--hf-output-len
|
|
--hf-split
|
|
--hf-subset
|
|
--host
|
|
--ignore-eos
|
|
--input-len
|
|
--insecure
|
|
--label
|
|
--logprobs
|
|
--lora-assignment
|
|
--lora-modules
|
|
--max-concurrency
|
|
--metadata
|
|
--metric-percentiles
|
|
--min-p
|
|
--model
|
|
--no-oversample
|
|
--no-reranker
|
|
--no-self-timed
|
|
--no-stream
|
|
--num-prompts
|
|
--num-warmups
|
|
--output-len
|
|
--percentile-metrics
|
|
--plot-dataset-stats
|
|
--plot-timeline
|
|
--port
|
|
--prefix-repetition-num-prefixes
|
|
--prefix-repetition-output-len
|
|
--prefix-repetition-prefix-len
|
|
--prefix-repetition-suffix-len
|
|
--presence-penalty
|
|
--probe-request-rate
|
|
--profile
|
|
--ramp-up-end-rps
|
|
--ramp-up-start-rps
|
|
--ramp-up-strategy
|
|
--random-batch-size
|
|
--random-input-len
|
|
--random-mm-base-items-per-request
|
|
--random-mm-bucket-config
|
|
--random-mm-limit-mm-per-prompt
|
|
--random-mm-num-mm-items-range-ratio
|
|
--random-output-len
|
|
--random-prefix-len
|
|
--random-range-ratio
|
|
--ready-check-timeout-sec
|
|
--repetition-penalty
|
|
--request-id-prefix
|
|
--request-rate
|
|
--result-dir
|
|
--result-filename
|
|
--save-detailed
|
|
--save-result
|
|
--seed
|
|
--self-timed
|
|
--served-model-name
|
|
--sharegpt-output-len
|
|
--skip-chat-template
|
|
--skip-tokenizer-init
|
|
--sonnet-input-len
|
|
--sonnet-output-len
|
|
--sonnet-prefix-len
|
|
--spec-bench-category
|
|
--spec-bench-output-len
|
|
--speed-bench-category
|
|
--speed-bench-dataset-subset
|
|
--speed-bench-output-len
|
|
--temperature
|
|
--timed-trace-chunk-hash-size
|
|
--timed-trace-label-hash-ids
|
|
--timed-trace-label-input-length
|
|
--timed-trace-label-output-length
|
|
--timed-trace-label-timestamp
|
|
--timed-trace-sec-multiplier
|
|
--timeline-itl-thresholds
|
|
--tokenizer
|
|
--tokenizer-mode
|
|
--top-k
|
|
--top-p
|
|
--trust-remote-code
|
|
--use-beam-search
|