KBaba7/llama.cpp

1### Server benchmark tools2 3Benchmark is using [k6](https://k6.io/).4 5##### Install k6 and sse extension6 7SSE is not supported by default in k6, you have to build k6 with the [xk6-sse](https://github.com/phymbert/xk6-sse) extension.8 9Example (assuming golang >= 1.21 is installed):10```shell11go install go.k6.io/xk6/cmd/xk6@latest12$GOPATH/bin/xk6 build master \13--with github.com/phymbert/xk6-sse14```15 16#### Download a dataset17 18This dataset was originally proposed in [vLLM benchmarks](https://github.com/vllm-project/vllm/blob/main/benchmarks/README.md).19 20```shell21wget https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/ShareGPT_V3_unfiltered_cleaned_split.json22```23 24#### Download a model25Example for PHI-226 27```shell28../../../scripts/hf.sh --repo ggml-org/models --file phi-2/ggml-model-q4_0.gguf29```30 31#### Start the server32The server must answer OAI Chat completion requests on `http://localhost:8080/v1` or according to the environment variable `SERVER_BENCH_URL`.33 34Example:35```shell36llama-server --host localhost --port 8080 \37  --model ggml-model-q4_0.gguf \38  --cont-batching \39  --metrics \40  --parallel 8 \41  --batch-size 512 \42  --ctx-size 4096 \43  -ngl 3344```45 46#### Run the benchmark47 48For 500 chat completions request with 8 concurrent users during maximum 10 minutes, run:49```shell50./k6 run script.js --duration 10m --iterations 500 --vus 851```52 53The benchmark values can be overridden with:54- `SERVER_BENCH_URL` server url prefix for chat completions, default `http://localhost:8080/v1`55- `SERVER_BENCH_N_PROMPTS` total prompts to randomly select in the benchmark, default `480`56- `SERVER_BENCH_MODEL_ALIAS` model alias to pass in the completion request, default `my-model`57- `SERVER_BENCH_MAX_TOKENS` max tokens to predict, default: `512`58- `SERVER_BENCH_DATASET` path to the benchmark dataset file59- `SERVER_BENCH_MAX_PROMPT_TOKENS` maximum prompt tokens to filter out in the dataset: default `1024`60- `SERVER_BENCH_MAX_CONTEXT` maximum context size of the completions request to filter out in the dataset: prompt + predicted tokens, default `2048`61 62Note: the local tokenizer is just a string space split, real number of tokens will differ.63 64Or with [k6 options](https://k6.io/docs/using-k6/k6-options/reference/):65 66```shell67SERVER_BENCH_N_PROMPTS=500 k6 run script.js --duration 10m --iterations 500 --vus 868```69 70To [debug http request](https://k6.io/docs/using-k6/http-debugging/) use `--http-debug="full"`.71 72#### Metrics73 74Following metrics are available computed from the OAI chat completions response `usage`:75- `llamacpp_tokens_second` Trend of `usage.total_tokens / request duration`76- `llamacpp_prompt_tokens` Trend of `usage.prompt_tokens`77- `llamacpp_prompt_tokens_total_counter` Counter of `usage.prompt_tokens`78- `llamacpp_completion_tokens` Trend of `usage.completion_tokens`79- `llamacpp_completion_tokens_total_counter` Counter of `usage.completion_tokens`80- `llamacpp_completions_truncated_rate` Rate of completions truncated, i.e. if `finish_reason === 'length'`81- `llamacpp_completions_stop_rate` Rate of completions stopped by the model, i.e. if `finish_reason === 'stop'`82 83The script will fail if too many completions are truncated, see `llamacpp_completions_truncated_rate`.84 85K6 metrics might be compared against [server metrics](../README.md), with:86 87```shell88curl http://localhost:8080/metrics89```90 91### Using the CI python script92The `bench.py` script does several steps:93- start the server94- define good variable for k695- run k6 script96- extract metrics from prometheus97 98It aims to be used in the CI, but you can run it manually:99 100```shell101LLAMA_SERVER_BIN_PATH=../../../cmake-build-release/bin/llama-server python bench.py \102              --runner-label local \103              --name local \104              --branch `git rev-parse --abbrev-ref HEAD` \105              --commit `git rev-parse HEAD` \106              --scenario script.js \107              --duration 5m \108              --hf-repo ggml-org/models	 \109              --hf-file phi-2/ggml-model-q4_0.gguf \110              --model-path-prefix models \111              --parallel 4 \112              -ngl 33 \113              --batch-size 2048 \114              --ubatch-size	256 \115              --ctx-size 4096 \116              --n-prompts 200 \117              --max-prompt-tokens 256 \118              --max-tokens 256119```120