search

vllm-project / vllm

A high-throughput and memory-efficient inference and serving engine for LLMs
https://docs.vllm.ai
Apache License 2.0
30.21k stars 4.57k forks source link

[Bug]: AttributeError: '_OpNamespace' '_C_cache_ops' object has no attribute 'reshape_and_cache' #9191

Open diegolovison opened 1 month ago

diegolovison commented 1 month ago

Your current environment

The output of `python collect_env.py` ```text (vllm-env) dlovison@dlovison:~/github/vllm$ python collect_env.py Collecting environment information... WARNING 10-09 09:50:35 _custom_ops.py:18] Failed to import from vllm._C with ModuleNotFoundError("No module named 'vllm._C'") /home/dlovison/github/vllm/vllm/connections.py:8: RuntimeWarning: Failed to read commit hash: No module named 'vllm._version' from vllm.version import __version__ as VLLM_VERSION 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: version 3.30.4 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: Could not collect CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA GeForce RTX 2060 Nvidia driver version: 550.107.02 cuDNN version: Could not collect 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): 20 On-line CPU(s) list: 0-19 Vendor ID: GenuineIntel Model name: 12th Gen Intel(R) Core(TM) i7-12700F CPU family: 6 Model: 151 Thread(s) per core: 2 Core(s) per socket: 12 Socket(s): 1 Stepping: 2 CPU max MHz: 4900,0000 CPU min MHz: 800,0000 BogoMIPS: 4224.00 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 epb cat_l2 cdp_l2 ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid rdt_a 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: 512 KiB (12 instances) L1i cache: 512 KiB (12 instances) L2 cache: 12 MiB (9 instances) L3 cache: 25 MiB (1 instance) NUMA node(s): 1 NUMA node0 CPU(s): 0-19 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==4.45.2 [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 4.45.2 pypi_0 pypi [conda] triton 3.0.0 pypi_0 pypi ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: dev vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X 0-19 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

Run benchmark_latency.py --device cuda from main branch

git log --oneline

a3691b6b (HEAD -> main, origin/main, origin/HEAD) [Core][Frontend] Add Support for Inference Time mm_processor_kwargs (#9131)
8c746226 [Frontend] API support for beam search for MQLLMEngine (#9117)
e1faa2a5 [misc] improve ux on readme (#9147)
80b57f00 [Intel GPU] Fix xpu decode input  (#9145)
04c12f81 [misc] update utils to support comparing multiple settings (#9140)
8eeb8570 Add Slack to README (#9137)

Stacktrace

/home/dlovison/miniconda3/envs/vllm-env/bin/python /home/dlovison/github/vllm/benchmarks/benchmark_latency.py --device cuda 
WARNING 10-09 09:48:06 _custom_ops.py:18] Failed to import from vllm._C with ModuleNotFoundError("No module named 'vllm._C'")
/home/dlovison/github/vllm/vllm/connections.py:8: RuntimeWarning: Failed to read commit hash:
No module named 'vllm._version'
  from vllm.version import __version__ as VLLM_VERSION
Namespace(model='facebook/opt-125m', speculative_model=None, num_speculative_tokens=None, speculative_draft_tensor_parallel_size=None, tokenizer=None, quantization=None, tensor_parallel_size=1, input_len=32, output_len=128, batch_size=8, n=1, use_beam_search=False, num_iters_warmup=10, num_iters=30, trust_remote_code=False, max_model_len=None, dtype='auto', enforce_eager=False, kv_cache_dtype='auto', quantization_param_path=None, profile=False, profile_result_dir=None, device='cuda', block_size=16, enable_chunked_prefill=False, enable_prefix_caching=False, use_v2_block_manager=False, ray_workers_use_nsight=False, download_dir=None, output_json=None, gpu_memory_utilization=0.9, load_format='auto', distributed_executor_backend=None, otlp_traces_endpoint=None)
INFO 10-09 09:48:12 llm_engine.py:237] Initializing an LLM engine (vdev) with config: model='facebook/opt-125m', speculative_config=None, tokenizer='facebook/opt-125m', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config=None, rope_scaling=None, rope_theta=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.float16, max_seq_len=2048, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=1, pipeline_parallel_size=1, disable_custom_all_reduce=False, quantization=None, enforce_eager=False, kv_cache_dtype=auto, quantization_param_path=None, device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='outlines'), observability_config=ObservabilityConfig(otlp_traces_endpoint=None, collect_model_forward_time=False, collect_model_execute_time=False), seed=0, served_model_name=facebook/opt-125m, use_v2_block_manager=False, num_scheduler_steps=1, chunked_prefill_enabled=False multi_step_stream_outputs=True, enable_prefix_caching=False, use_async_output_proc=True, use_cached_outputs=False, mm_processor_kwargs=None)
INFO 10-09 09:48:13 selector.py:217] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
INFO 10-09 09:48:13 selector.py:116] Using XFormers backend.
/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/xformers/ops/fmha/flash.py:211: FutureWarning: `torch.library.impl_abstract` was renamed to `torch.library.register_fake`. Please use that instead; we will remove `torch.library.impl_abstract` in a future version of PyTorch.
  @torch.library.impl_abstract("xformers_flash::flash_fwd")
/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/xformers/ops/fmha/flash.py:344: FutureWarning: `torch.library.impl_abstract` was renamed to `torch.library.register_fake`. Please use that instead; we will remove `torch.library.impl_abstract` in a future version of PyTorch.
  @torch.library.impl_abstract("xformers_flash::flash_bwd")
INFO 10-09 09:48:13 model_runner.py:1051] Starting to load model facebook/opt-125m...
INFO 10-09 09:48:13 selector.py:217] Cannot use FlashAttention-2 backend for Volta and Turing GPUs.
INFO 10-09 09:48:13 selector.py:116] Using XFormers backend.
INFO 10-09 09:48:13 weight_utils.py:242] Using model weights format ['*.bin']
Loading pt checkpoint shards:   0% Completed | 0/1 [00:00<?, ?it/s]
/home/dlovison/github/vllm/vllm/model_executor/model_loader/weight_utils.py:424: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
  state = torch.load(bin_file, map_location="cpu")
Loading pt checkpoint shards: 100% Completed | 1/1 [00:00<00:00,  9.13it/s]
Loading pt checkpoint shards: 100% Completed | 1/1 [00:00<00:00,  9.12it/s]

INFO 10-09 09:48:20 model_runner.py:1062] Loading model weights took 0.2389 GB
INFO 10-09 09:48:21 gpu_executor.py:122] # GPU blocks: 8017, # CPU blocks: 7281
INFO 10-09 09:48:22 model_runner.py:1385] Capturing the model for CUDA graphs. This may lead to unexpected consequences if the model is not static. To run the model in eager mode, set 'enforce_eager=True' or use '--enforce-eager' in the CLI.
INFO 10-09 09:48:22 model_runner.py:1389] CUDA graphs can take additional 1~3 GiB memory per GPU. If you are running out of memory, consider decreasing `gpu_memory_utilization` or enforcing eager mode. You can also reduce the `max_num_seqs` as needed to decrease memory usage.
ERROR 10-09 09:48:22 _custom_ops.py:51] Error in calling custom op reshape_and_cache: '_OpNamespace' '_C_cache_ops' object has no attribute 'reshape_and_cache'
ERROR 10-09 09:48:22 _custom_ops.py:51] Possibly you have built or installed an obsolete version of vllm.
ERROR 10-09 09:48:22 _custom_ops.py:51] Please try a clean build and install of vllm,or remove old built files such as vllm/*cpython*.so and build/ .
[rank0]: Traceback (most recent call last):
[rank0]:   File "/home/dlovison/github/vllm/benchmarks/benchmark_latency.py", line 282, in <module>
[rank0]:     main(args)
[rank0]:   File "/home/dlovison/github/vllm/benchmarks/benchmark_latency.py", line 24, in main
[rank0]:     llm = LLM(
[rank0]:           ^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/entrypoints/llm.py", line 177, in __init__
[rank0]:     self.llm_engine = LLMEngine.from_engine_args(
[rank0]:                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/engine/llm_engine.py", line 574, in from_engine_args
[rank0]:     engine = cls(
[rank0]:              ^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/engine/llm_engine.py", line 349, in __init__
[rank0]:     self._initialize_kv_caches()
[rank0]:   File "/home/dlovison/github/vllm/vllm/engine/llm_engine.py", line 497, in _initialize_kv_caches
[rank0]:     self.model_executor.initialize_cache(num_gpu_blocks, num_cpu_blocks)
[rank0]:   File "/home/dlovison/github/vllm/vllm/executor/gpu_executor.py", line 125, in initialize_cache
[rank0]:     self.driver_worker.initialize_cache(num_gpu_blocks, num_cpu_blocks)
[rank0]:   File "/home/dlovison/github/vllm/vllm/worker/worker.py", line 266, in initialize_cache
[rank0]:     self._warm_up_model()
[rank0]:   File "/home/dlovison/github/vllm/vllm/worker/worker.py", line 282, in _warm_up_model
[rank0]:     self.model_runner.capture_model(self.gpu_cache)
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
[rank0]:     return func(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/worker/model_runner.py", line 1505, in capture_model
[rank0]:     graph_runner.capture(**capture_inputs)
[rank0]:   File "/home/dlovison/github/vllm/vllm/worker/model_runner.py", line 1768, in capture
[rank0]:     self.model(
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/model_executor/models/opt.py", line 346, in forward
[rank0]:     hidden_states = self.model(input_ids, positions, kv_caches,
[rank0]:                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/model_executor/models/opt.py", line 308, in forward
[rank0]:     return self.decoder(input_ids,
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/model_executor/models/opt.py", line 269, in forward
[rank0]:     hidden_states = layer(hidden_states,
[rank0]:                     ^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/model_executor/models/opt.py", line 165, in forward
[rank0]:     hidden_states = self.self_attn(hidden_states=hidden_states,
[rank0]:                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/model_executor/models/opt.py", line 108, in forward
[rank0]:     attn_output = self.attn(q, k, v, kv_cache, attn_metadata)
[rank0]:                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
[rank0]:     return self._call_impl(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
[rank0]:     return forward_call(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/attention/layer.py", line 98, in forward
[rank0]:     return self.impl.forward(query,
[rank0]:            ^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/attention/backends/xformers.py", line 559, in forward
[rank0]:     PagedAttention.write_to_paged_cache(key, value, key_cache,
[rank0]:   File "/home/dlovison/github/vllm/vllm/attention/ops/paged_attn.py", line 75, in write_to_paged_cache
[rank0]:     ops.reshape_and_cache(
[rank0]:   File "/home/dlovison/github/vllm/vllm/_custom_ops.py", line 52, in wrapper
[rank0]:     raise e
[rank0]:   File "/home/dlovison/github/vllm/vllm/_custom_ops.py", line 34, in wrapper
[rank0]:     return fn(*args, **kwargs)
[rank0]:            ^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/github/vllm/vllm/_custom_ops.py", line 867, in reshape_and_cache
[rank0]:     torch.ops._C_cache_ops.reshape_and_cache(key, value, key_cache,
[rank0]:     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[rank0]:   File "/home/dlovison/miniconda3/envs/vllm-env/lib/python3.11/site-packages/torch/_ops.py", line 1170, in __getattr__
[rank0]:     raise AttributeError(
[rank0]: AttributeError: '_OpNamespace' '_C_cache_ops' object has no attribute 'reshape_and_cache'

Process finished with exit code 1

Before submitting a new issue...

sathvikswaminathan commented 1 day ago

Facing the exact same issue on CPUs, tried doing a fresh install and compiling from source, error still persists.