search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
30.42k stars 4.6k forks source link

[Bug]: Continuous usage stats are incorrect when chunked prefill is enabled #9028

Open tdoublep opened 1 month ago

tdoublep commented 1 month ago

Your current environment

The output of `python collect_env.py` ```text $ python collect_env.py Collecting environment information... PyTorch version: 2.4.0+cu121 Is debug build: False CUDA used to build PyTorch: 12.1 ROCM used to build PyTorch: N/A OS: Ubuntu 22.04.4 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: Could not collect CMake version: version 3.30.3 Libc version: glibc-2.35 Python version: 3.11.9 | packaged by conda-forge | (main, Apr 19 2024, 18:36:13) [GCC 12.3.0] (64-bit runtime) Python platform: Linux-5.15.0-101-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: Could not collect CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA H100 80GB HBM3 GPU 1: NVIDIA H100 80GB HBM3 GPU 2: NVIDIA H100 80GB HBM3 GPU 3: NVIDIA H100 80GB HBM3 GPU 4: NVIDIA H100 80GB HBM3 GPU 5: NVIDIA H100 80GB HBM3 GPU 6: NVIDIA H100 80GB HBM3 GPU 7: NVIDIA H100 80GB HBM3 Nvidia driver version: 550.54.15 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.9.7 /usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.9.7 HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True CPU: Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 52 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 192 On-line CPU(s) list: 0-191 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) Platinum 8474C CPU family: 6 Model: 143 Thread(s) per core: 2 Core(s) per socket: 48 Socket(s): 2 Stepping: 8 CPU max MHz: 3800.0000 CPU min MHz: 800.0000 BogoMIPS: 4200.00 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb cat_l3 cat_l2 cdp_l3 invpcid_single intel_ppin cdp_l2 ssbd mba ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb intel_pt avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local split_lock_detect avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts hwp hwp_act_window hwp_epp hwp_pkg_req avx512vbmi umip pku ospke waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid bus_lock_detect cldemote movdiri movdir64b enqcmd fsrm md_clear serialize tsxldtrk pconfig arch_lbr amx_bf16 avx512_fp16 amx_tile amx_int8 flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 4.5 MiB (96 instances) L1i cache: 3 MiB (96 instances) L2 cache: 192 MiB (96 instances) L3 cache: 195 MiB (2 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-47,96-143 NUMA node1 CPU(s): 48-95,144-191 Vulnerability Gather data sampling: Not affected Vulnerability Itlb multihit: Not affected Vulnerability L1tf: Not affected Vulnerability Mds: Not affected Vulnerability Meltdown: Not affected Vulnerability Mmio stale data: Not affected Vulnerability Retbleed: Not affected Vulnerability Spec rstack overflow: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl and seccomp Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence Vulnerability Srbds: Not affected Vulnerability Tsx async abort: Not affected Versions of relevant libraries: [pip3] numpy==1.26.4 [pip3] nvidia-cublas-cu12==12.1.3.1 [pip3] nvidia-cuda-cupti-cu12==12.1.105 [pip3] nvidia-cuda-nvrtc-cu12==12.1.105 [pip3] nvidia-cuda-runtime-cu12==12.1.105 [pip3] nvidia-cudnn-cu12==9.1.0.70 [pip3] nvidia-cufft-cu12==11.0.2.54 [pip3] nvidia-curand-cu12==10.3.2.106 [pip3] nvidia-cusolver-cu12==11.4.5.107 [pip3] nvidia-cusparse-cu12==12.1.0.106 [pip3] nvidia-ml-py==12.560.30 [pip3] nvidia-nccl-cu12==2.20.5 [pip3] nvidia-nvjitlink-cu12==12.6.68 [pip3] nvidia-nvtx-cu12==12.1.105 [pip3] pyzmq==26.2.0 [pip3] torch==2.4.0 [pip3] torchvision==0.19.0 [pip3] transformers==4.45.1 [pip3] triton==3.0.0 [conda] numpy 1.26.4 pypi_0 pypi [conda] nvidia-cublas-cu12 12.1.3.1 pypi_0 pypi [conda] nvidia-cuda-cupti-cu12 12.1.105 pypi_0 pypi [conda] nvidia-cuda-nvrtc-cu12 12.1.105 pypi_0 pypi [conda] nvidia-cuda-runtime-cu12 12.1.105 pypi_0 pypi [conda] nvidia-cudnn-cu12 9.1.0.70 pypi_0 pypi [conda] nvidia-cufft-cu12 11.0.2.54 pypi_0 pypi [conda] nvidia-curand-cu12 10.3.2.106 pypi_0 pypi [conda] nvidia-cusolver-cu12 11.4.5.107 pypi_0 pypi [conda] nvidia-cusparse-cu12 12.1.0.106 pypi_0 pypi [conda] nvidia-ml-py 12.560.30 pypi_0 pypi [conda] nvidia-nccl-cu12 2.20.5 pypi_0 pypi [conda] nvidia-nvjitlink-cu12 12.6.68 pypi_0 pypi [conda] nvidia-nvtx-cu12 12.1.105 pypi_0 pypi [conda] pyzmq 26.2.0 pypi_0 pypi [conda] torch 2.4.0 pypi_0 pypi [conda] torchvision 0.19.0 pypi_0 pypi [conda] transformers 4.45.1 pypi_0 pypi [conda] triton 3.0.0 pypi_0 pypi ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: N/A vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 NIC8 NIC9 NIC10 NIC11 NIC12 NIC13 NIC14 NIC15 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS 0-47,96-143 N/A GPU1 SYS X SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS 0-47,96-143 N/A GPU2 SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS 0-47,96-143 N/A GPU3 SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS SYS SYS SYS SYS 0-47,96-143 N/A GPU4 SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS SYS SYS 48-95,144-19N/A GPU5 SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS SYS SYS 48-95,144-19N/A GPU6 SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX SYS SYS 48-95,144-19N/A GPU7 SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX PIX 48-95,144-19N/A NIC0 PIX SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC1 PIX SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC2 SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC3 SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC4 SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC5 SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS NIC6 SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS SYS SYS SYS SYS NIC7 SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS SYS SYS SYS SYS NIC8 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS SYS SYS NIC9 SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS SYS SYS NIC10 SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS SYS SYS NIC11 SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS SYS SYS NIC12 SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX SYS SYS NIC13 SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X SYS SYS NIC14 SYS SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS X PIX NIC15 SYS SYS SYS SYS SYS SYS SYS PIX SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS SYS PIX X Legend: X = Self SYS = Connection traversing PCIe as well as the SMP interconnect between NUMA nodes (e.g., QPI/UPI) NODE = Connection traversing PCIe as well as the interconnect between PCIe Host Bridges within a NUMA node PHB = Connection traversing PCIe as well as a PCIe Host Bridge (typically the CPU) PXB = Connection traversing multiple PCIe bridges (without traversing the PCIe Host Bridge) PIX = Connection traversing at most a single PCIe bridge NV# = Connection traversing a bonded set of # NVLinks NIC Legend: NIC0: mlx5_0 NIC1: mlx5_1 NIC2: mlx5_2 NIC3: mlx5_3 NIC4: mlx5_4 NIC5: mlx5_5 NIC6: mlx5_6 NIC7: mlx5_7 NIC8: mlx5_8 NIC9: mlx5_9 NIC10: mlx5_10 NIC11: mlx5_11 NIC12: mlx5_12 NIC13: mlx5_13 NIC14: mlx5_14 NIC15: mlx5_15 ```

Model Input Dumps

No response

🐛 Describe the bug

Start the inference server with:

python3 -m vllm.entrypoints.openai.api_server --enable-chunked-prefill

Then send a request with a long prompt for a single output token and enable streaming usage stats:

import requests
import json

openai_api_base = "http://localhost:8000/v1"

model = requests.get("%s/models" % (openai_api_base)).json()["data"][0]["id"]

prompt = "test " * 1_000

request = {
    "model": model,
    "prompt": prompt,
    "max_tokens": 1,
    "temperature": 0,
    "stream": True,
    "stream_options": {"include_usage": True, "continuous_usage_stats": True},
}

headers = {"User-Agent": "Test Client"}

response = requests.post(
    "%s/completions" % (openai_api_base),
    headers=headers,
    json=request,
    stream=True,
)

finished = False
for chunk in response.iter_lines(
    chunk_size=8192, decode_unicode=False, delimiter=b"\n"
):
    if chunk and not finished:
        data = chunk.decode("utf-8").strip().split("data: ")[1]
        data_parsed = json.loads(data)
        print(json.dumps(data_parsed, indent=2))
        finished = data_parsed["choices"][0]["finish_reason"] is not None

produces:

{
  "id": "cmpl-fa1dcd08905f497c9cf514eba80b48da",
  "object": "text_completion",
  "created": 1727900139,
  "model": "facebook/opt-125m",
  "choices": [
    {
      "index": 0,
      "text": "",
      "logprobs": null,
      "finish_reason": null,
      "stop_reason": null
    }
  ],
  "usage": {
    "prompt_tokens": 1002,
    "total_tokens": 2004,
    "completion_tokens": 1002
  }
}
{
  "id": "cmpl-fa1dcd08905f497c9cf514eba80b48da",
  "object": "text_completion",
  "created": 1727900139,
  "model": "facebook/opt-125m",
  "choices": [
    {
      "index": 0,
      "text": " test",
      "logprobs": null,
      "finish_reason": "length",
      "stop_reason": null
    }
  ],
  "usage": {
    "prompt_tokens": 1002,
    "total_tokens": 2005,
    "completion_tokens": 1003
  }
}

The token counting is totally wrong. The first response should have completion_tokens=0, this is how it behaved in previous versions of vLLM. Works fine without chunked prefill.

Might be related to #8625 but looks slightly different? I will debug it now and take a look at that one too.

Before submitting a new issue...

tdoublep commented 1 month ago

@njhill I saw you cleaned up this code recently. Did you happen to check the case with chunked prefill too? It looked like it was broken a couple of weeks ago.

njhill commented 3 weeks ago

@tdoublep I expect the recent changes I made fixed this issue, they included skipping streaming responses for intermediate prompt chunks. I will verify that when I get a chance!