search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
31.17k stars 4.74k forks source link

[Bug]: Unsupported base layer: QKVParallelLinear when loading lora to a quantized model #9120

Closed fahadh4ilyas closed 1 month ago

fahadh4ilyas commented 1 month ago

Your current environment

The output of `python collect_env.py` ```text 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 22.04.5 LTS (x86_64) GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 Clang version: Could not collect CMake version: Could not collect Libc version: glibc-2.35 Python version: 3.11.10 (main, Oct 3 2024, 07:29:13) [GCC 11.2.0] (64-bit runtime) Python platform: Linux-6.8.0-45-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: 12.2.140 CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA GeForce RTX 4090 GPU 1: NVIDIA GeForce RTX 4090 Nvidia driver version: 535.183.06 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_adv_infer.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_adv_train.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_infer.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_cnn_train.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_ops_infer.so.8.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_ops_train.so.8.5.0 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: 39 bits physical, 48 bits virtual Byte Order: Little Endian CPU(s): 16 On-line CPU(s) list: 0-15 Vendor ID: GenuineIntel Model name: 13th Gen Intel(R) Core(TM) i5-13400 CPU family: 6 Model: 191 Thread(s) per core: 2 Core(s) per socket: 10 Socket(s): 1 Stepping: 2 CPU max MHz: 4600,0000 CPU min MHz: 800,0000 BogoMIPS: 4992.00 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 est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rdseed adx smap clflushopt clwb intel_pt sha_ni xsaveopt xsavec xgetbv1 xsaves split_lock_detect user_shstk avx_vnni dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp hwp_pkg_req hfi vnmi umip pku ospke waitpkg gfni vaes vpclmulqdq rdpid movdiri movdir64b fsrm md_clear serialize arch_lbr ibt flush_l1d arch_capabilities Virtualization: VT-x L1d cache: 416 KiB (10 instances) L1i cache: 448 KiB (10 instances) L2 cache: 9,5 MiB (7 instances) L3 cache: 20 MiB (1 instance) NUMA node(s): 1 NUMA node0 CPU(s): 0-15 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 Reg file data sampling: Mitigation; Clear Register File Vulnerability Retbleed: Not affected Vulnerability Spec rstack overflow: Not affected Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization Vulnerability Spectre v2: Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S 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==12.6.77 [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 12.6.77 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 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X PHB 0-15 0 N/A GPU1 PHB X 0-15 0 N/A 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 ```

Model Input Dumps

No response

🐛 Describe the bug

Here is how I load my model using vllm serve

vllm serve <my-model-path> --host 0.0.0.0 --served-model-name my-model --port 8000 --enable-lora --lora-modules my-model:lora=<my-lora-path>

And here is the error that I got

Process SpawnProcess-1:
Traceback (most recent call last):
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/multiprocessing/process.py", line 314, in _bootstrap
    self.run()
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/home/fahadh/vllm/vllm/engine/multiprocessing/engine.py", line 388, in run_mp_engine
    engine = MQLLMEngine.from_engine_args(engine_args=engine_args,
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/engine/multiprocessing/engine.py", line 138, in from_engine_args
    return cls(
           ^^^^
  File "/home/fahadh/vllm/vllm/engine/multiprocessing/engine.py", line 78, in __init__
    self.engine = LLMEngine(*args,
                  ^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/engine/llm_engine.py", line 338, in __init__
    self.model_executor = executor_class(
                          ^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/executor/executor_base.py", line 47, in __init__
    self._init_executor()
  File "/home/fahadh/vllm/vllm/executor/gpu_executor.py", line 40, in _init_executor
    self.driver_worker.load_model()
  File "/home/fahadh/vllm/vllm/worker/worker.py", line 183, in load_model
    self.model_runner.load_model()
  File "/home/fahadh/vllm/vllm/worker/model_runner.py", line 1045, in load_model
    self.model = self.lora_manager.create_lora_manager(self.model)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/lora/worker_manager.py", line 173, in create_lora_manager
    lora_manager = create_lora_manager(
                   ^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/lora/models.py", line 734, in create_lora_manager
    lora_manager = lora_manager_cls(
                   ^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/lora/models.py", line 658, in __init__
    super().__init__(model, max_num_seqs, max_num_batched_tokens,
  File "/home/fahadh/vllm/vllm/lora/models.py", line 343, in __init__
    self._create_lora_modules()
  File "/home/fahadh/vllm/vllm/lora/models.py", line 457, in _create_lora_modules
    from_layer(module, self.lora_slots, self.lora_config,
  File "/home/fahadh/vllm/vllm/lora/utils.py", line 64, in from_layer
    ret = lora_cls(layer)
          ^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/lora/layers.py", line 694, in __init__
    super().__init__(base_layer)
  File "/home/fahadh/vllm/vllm/lora/layers.py", line 372, in __init__
    self.device = _get_lora_device(self.base_layer)
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/lora/layers.py", line 49, in _get_lora_device
    raise ValueError(f"Unsupported base layer: {base_layer}")
ValueError: Unsupported base layer: QKVParallelLinear(in_features=4096, output_features=6144, bias=False, tp_size=1, gather_output=False)
Traceback (most recent call last):
  File "/home/fahadh/anaconda3/envs/vllm/bin/vllm", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/fahadh/vllm/vllm/scripts.py", line 191, in main
    args.dispatch_function(args)
  File "/home/fahadh/vllm/vllm/scripts.py", line 40, in serve
    uvloop.run(run_server(args))
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/site-packages/uvloop/__init__.py", line 105, in run
    return runner.run(wrapper())
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "uvloop/loop.pyx", line 1517, in uvloop.loop.Loop.run_until_complete
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/site-packages/uvloop/__init__.py", line 61, in wrapper
    return await main
           ^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/entrypoints/openai/api_server.py", line 538, in run_server
    async with build_async_engine_client(args) as engine_client:
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/entrypoints/openai/api_server.py", line 105, in build_async_engine_client
    async with build_async_engine_client_from_engine_args(
  File "/home/fahadh/anaconda3/envs/vllm/lib/python3.11/contextlib.py", line 210, in __aenter__
    return await anext(self.gen)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/fahadh/vllm/vllm/entrypoints/openai/api_server.py", line 192, in build_async_engine_client_from_engine_args
    raise RuntimeError(
RuntimeError: Engine process failed to start

Here is what inside config.json of my model

{
  "_name_or_path": "my-model",
  "architectures": [
    "MistralForCausalLM"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 32000,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 14336,
  "max_position_embeddings": 8192,
  "model_type": "mistral",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "num_key_value_heads": 8,
  "quantization_config": {
    "config_groups": {
      "group_0": {
        "input_activations": null,
        "output_activations": null,
        "targets": [
          "Linear"
        ],
        "weights": {
          "actorder": null,
          "block_structure": null,
          "dynamic": false,
          "group_size": null,
          "num_bits": 8,
          "observer": "minmax",
          "observer_kwargs": {},
          "strategy": "channel",
          "symmetric": true,
          "type": "int"
        }
      }
    },
    "format": "pack-quantized",
    "global_compression_ratio": 1.460841389791697,
    "ignore": [
      "lm_head"
    ],
    "kv_cache_scheme": null,
    "quant_method": "compressed-tensors",
    "quantization_status": "compressed"
  },
  "rms_norm_eps": 1e-05,
  "rope_theta": 10000.0,
  "sliding_window": 4096,
  "tie_word_embeddings": false,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.45.1",
  "use_cache": true,
  "vocab_size": 32019
}

Before submitting a new issue...

jeejeelee commented 1 month ago

it seems that this error raised by using compressed-tensors. It indeed is a bug, i will inviestgaet it asap.

fahadh4ilyas commented 1 month ago

it seems that this error raised by using compressed-tensors. It indeed is a bug, i will inviestgaet it asap.

I found the solution. Just change line 41 from vllm/lora/layers.py from

        return base_layer.weight.device

to

        return base_layer.weight.device
    elif hasattr(base_layer, "weight_packed"):
        return base_layer.weight_packed.device

Because it seems that compressed-tensors are using weight_packed instead of weight to store parameters