[Bug]: There is an issue with speculative inference in Eagle mode, where the context length of vLLM inference is constrained by the draft model. #21986
Description
Your current environment
Collecting environment information...
System Info
==============================
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 4.0.2
Libc version : glibc-2.35
==============================
PyTorch Info
PyTorch version : 2.7.1+cu126
Is debug build : False
CUDA used to build PyTorch : 12.6
ROCM used to build PyTorch : N/A
==============================
Python Environment
Python version : 3.12.10 (main, Apr 9 2025, 08:55:05) [GCC 11.4.0] (64-bit runtime)
Python platform : Linux-5.15.0-50-generic-x86_64-with-glibc2.35
==============================
CUDA / GPU Info
Is CUDA available : True
CUDA runtime version : 12.8.93
CUDA_MODULE_LOADING set to : LAZY
GPU models and configuration : GPU 0: NVIDIA H20
==============================
CPU Info
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 208
On-line CPU(s) list: 0-207
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 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 avx_vnni avx512_bf16 wbnoinvd dtherm ida arat pln pts 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: 4.9 MiB (104 instances)
L1i cache: 3.3 MiB (104 instances)
L2 cache: 208 MiB (104 instances)
L3 cache: 640 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-51,104-155
NUMA node1 CPU(s): 52-103,156-207
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 store bypass: Vulnerable
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2: Vulnerable, IBPB: disabled, STIBP: disabled, PBRSB-eIBRS: Vulnerable
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
==============================
Versions of relevant libraries
[pip3] numpy==2.2.6
[pip3] nvidia-cublas-cu12==12.6.4.1
[pip3] nvidia-cuda-cupti-cu12==12.6.80
[pip3] nvidia-cuda-nvrtc-cu12==12.6.77
[pip3] nvidia-cuda-runtime-cu12==12.6.77
[pip3] nvidia-cudnn-cu12==9.5.1.17
[pip3] nvidia-cufft-cu12==11.3.0.4
[pip3] nvidia-cufile-cu12==1.11.1.6
[pip3] nvidia-curand-cu12==10.3.7.77
[pip3] nvidia-cusolver-cu12==11.7.1.2
[pip3] nvidia-cusparse-cu12==12.5.4.2
[pip3] nvidia-cusparselt-cu12==0.6.3
[pip3] nvidia-nccl-cu12==2.26.2
[pip3] nvidia-nvjitlink-cu12==12.6.85
[pip3] nvidia-nvtx-cu12==12.6.77
[pip3] pyzmq==26.4.0
[pip3] torch==2.7.1
[pip3] torchaudio==2.7.1
[pip3] torchvision==0.22.1
[pip3] transformers==4.54.1
[pip3] triton==3.3.1
[conda] Could not collect
==============================
vLLM Info
ROCM Version : Could not collect
Neuron SDK Version : N/A
vLLM Version : 0.10.0
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
GPU0 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 NIC7 NIC8 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NODE PIX NODE NODE NODE SYS SYS SYS SYS 0-51,104-155 0 N/A
NIC0 NODE X NODE NODE NODE NODE SYS SYS SYS SYS
==============================
Environment Variables
NVIDIA_VISIBLE_DEVICES=all
NVIDIA_REQUIRE_CUDA=cuda>=12.8 brand=unknown,driver>=470,driver<471 brand=grid,driver>=470,driver<471 brand=tesla,driver>=470,driver<471 brand=nvidia,driver>=470,driver<471 brand=quadro,driver>=470,driver<471 brand=quadrortx,driver>=470,driver<471 brand=nvidiartx,driver>=470,driver<471 brand=vapps,driver>=470,driver<471 brand=vpc,driver>=470,driver<471 brand=vcs,driver>=470,driver<471 brand=vws,driver>=470,driver<471 brand=cloudgaming,driver>=470,driver<471 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=560,driver<561 brand=grid,driver>=560,driver<561 brand=tesla,driver>=560,driver<561 brand=nvidia,driver>=560,driver<561 brand=quadro,driver>=560,driver<561 brand=quadrortx,driver>=560,driver<561 brand=nvidiartx,driver>=560,driver<561 brand=vapps,driver>=560,driver<561 brand=vpc,driver>=560,driver<561 brand=vcs,driver>=560,driver<561 brand=vws,driver>=560,driver<561 brand=cloudgaming,driver>=560,driver<561 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566
NCCL_VERSION=2.25.1-1
NVIDIA_DRIVER_CAPABILITIES=compute,utility
NVIDIA_PRODUCT_NAME=CUDA
VLLM_USAGE_SOURCE=production-docker-image
CUDA_VERSION=12.8.1
LD_LIBRARY_PATH=/usr/local/cuda/lib64
NCCL_CUMEM_ENABLE=0
PYTORCH_NVML_BASED_CUDA_CHECK=1
TORCHINDUCTOR_COMPILE_THREADS=1
CUDA_MODULE_LOADING=LAZY
🐛 Describe the bug
There is an issue with speculative inference in Eagle mode, where the context length of vLLM inference is constrained by the draft model. For example, the draft model "EAGLE3-LLaMA3.1-Instruct-8B" has a "max_position_embeddings" limit of 2048, which affects the inference context length.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [81,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [82,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [83,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [84,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [85,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [86,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/inductor_cache/bm/cbmyyoelcvytx6fwgno32vdhljvglspjh6ltrcpur2q44wg6a42j.py:37: unknown: block: [9,0,0], thread: [87,0,0] Assertion index out of bounds: 0 <= tl.broadcast_to(tmp10, [XBLOCK]) < 2048 failed.
ERROR 07-30 20:29:47 [dump_input.py:69] Dumping input data for V1 LLM engine (v0.10.0) with config: model='/model/Llama3.1-8B-Instruct', speculative_config=SpeculativeConfig(method='eagle3', model='/model/EAGLE3-LLaMA3.1-Instruct-8B', num_spec_tokens=2), tokenizer='/model/Llama3.1-8B-Instruct', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config={}, tokenizer_revision=None, trust_remote_code=False, dtype=torch.float16, max_seq_len=65536, 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, device_config=cuda, decoding_config=DecodingConfig(backend='auto', disable_fallback=False, disable_any_whitespace=False, disable_additional_properties=False, reasoning_backend=''), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None), seed=0, served_model_name=Meta-Llama-3.1-8B-Instruct, num_scheduler_steps=1, multi_step_stream_outputs=True, enable_prefix_caching=True, chunked_prefill_enabled=True, use_async_output_proc=False, pooler_config=None, compilation_config={"level":3,"debug_dump_path":"","cache_dir":"/root/.cache/vllm/torch_compile_cache/6a2e0f978f","backend":"","custom_ops":[],"splitting_ops":["vllm.unified_attention","vllm.unified_attention_with_output","vllm.mamba_mixer2"],"use_inductor":true,"compile_sizes":[],"inductor_compile_config":{"enable_auto_functionalized_v2":false},"inductor_passes":{},"use_cudagraph":true,"cudagraph_num_of_warmups":1,"cudagraph_capture_sizes":[512,504,496,488,480,472,464,456,448,440,432,424,416,408,400,392,384,376,368,360,352,344,336,328,320,312,304,296,288,280,272,264,256,248,240,232,224,216,208,200,192,184,176,168,160,152,144,136,128,120,112,104,96,88,80,72,64,56,48,40,32,24,16,8,4,2,1],"cudagraph_copy_inputs":false,"full_cuda_graph":false,"max_capture_size":512,"local_cache_dir":"/root/.cache/vllm/torch_compile_cache/6a2e0f978f/rank_0_0/eagle_head"},
ERROR 07-30 20:29:47 [dump_input.py:76] Dumping scheduler output for model execution: SchedulerOutput(scheduled_new_reqs=[], scheduled_cached_reqs=CachedRequestData(req_ids=['chatcmpl-698a2381c5aa43e290b3d7bd358817a1'], resumed_from_preemption=[false], new_token_ids=[], new_block_ids=[[[]]], num_computed_tokens=[2047]), num_scheduled_tokens={chatcmpl-698a2381c5aa43e290b3d7bd358817a1: 3}, total_num_scheduled_tokens=3, scheduled_spec_decode_tokens={chatcmpl-698a2381c5aa43e290b3d7bd358817a1: [499, 2351]}, scheduled_encoder_inputs={}, num_common_prefix_blocks=[129], finished_req_ids=[], free_encoder_input_ids=[], structured_output_request_ids={}, grammar_bitmask=null, kv_connector_metadata=null)
ERROR 07-30 20:29:47 [dump_input.py:79] Dumping scheduler stats: SchedulerStats(num_running_reqs=1, num_waiting_reqs=0, kv_cache_usage=0.0037155596204413444, prefix_cache_stats=PrefixCacheStats(reset=False, requests=0, queries=0, hits=0), spec_decoding_stats=None, num_corrupted_reqs=0)
ERROR 07-30 20:29:47 [core.py:634] EngineCore encountered a fatal error.
ERROR 07-30 20:29:47 [core.py:634] Traceback (most recent call last):
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 625, in run_engine_core
ERROR 07-30 20:29:47 [core.py:634] engine_core.run_busy_loop()
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 652, in run_busy_loop
ERROR 07-30 20:29:47 [core.py:634] self._process_engine_step()
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 677, in _process_engine_step
ERROR 07-30 20:29:47 [core.py:634] outputs, model_executed = self.step_fn()
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 267, in step
ERROR 07-30 20:29:47 [core.py:634] model_output = self.execute_model_with_error_logging(
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 253, in execute_model_with_error_logging
ERROR 07-30 20:29:47 [core.py:634] raise err
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 244, in execute_model_with_error_logging
ERROR 07-30 20:29:47 [core.py:634] return model_fn(scheduler_output)
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/executor/abstract.py", line 87, in execute_model
ERROR 07-30 20:29:47 [core.py:634] output = self.collective_rpc("execute_model",
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/executor/uniproc_executor.py", line 58, in collective_rpc
ERROR 07-30 20:29:47 [core.py:634] answer = run_method(self.driver_worker, method, args, kwargs)
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/utils/init.py", line 2985, in run_method
ERROR 07-30 20:29:47 [core.py:634] return func(*args, **kwargs)
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/torch/utils/_contextlib.py", line 116, in decorate_context
ERROR 07-30 20:29:47 [core.py:634] return func(*args, **kwargs)
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_worker.py", line 337, in execute_model
ERROR 07-30 20:29:47 [core.py:634] output = self.model_runner.execute_model(scheduler_output,
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/torch/utils/_contextlib.py", line 116, in decorate_context
ERROR 07-30 20:29:47 [core.py:634] return func(*args, **kwargs)
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_model_runner.py", line 1622, in execute_model
ERROR 07-30 20:29:47 [core.py:634] spec_token_ids = self.propose_draft_token_ids(
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_model_runner.py", line 1745, in propose_draft_token_ids
ERROR 07-30 20:29:47 [core.py:634] spec_token_ids = draft_token_ids.tolist()
ERROR 07-30 20:29:47 [core.py:634] ^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [core.py:634] RuntimeError: CUDA error: device-side assert triggered
ERROR 07-30 20:29:47 [core.py:634] CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
ERROR 07-30 20:29:47 [core.py:634] For debugging consider passing CUDA_LAUNCH_BLOCKING=1
ERROR 07-30 20:29:47 [core.py:634] Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
ERROR 07-30 20:29:47 [core.py:634]
Process EngineCore_0:
Traceback (most recent call last):
File "/usr/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
self.run()
File "/usr/lib/python3.12/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 636, in run_engine_core
raise e
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 625, in run_engine_core
engine_core.run_busy_loop()
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 652, in run_busy_loop
self._process_engine_step()
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 677, in _process_engine_step
outputs, model_executed = self.step_fn()
^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 267, in step
model_output = self.execute_model_with_error_logging(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 253, in execute_model_with_error_logging
raise err
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core.py", line 244, in execute_model_with_error_logging
return model_fn(scheduler_output)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/executor/abstract.py", line 87, in execute_model
output = self.collective_rpc("execute_model",
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/executor/uniproc_executor.py", line 58, in collective_rpc
answer = run_method(self.driver_worker, method, args, kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/utils/init.py", line 2985, in run_method
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/torch/utils/_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_worker.py", line 337, in execute_model
output = self.model_runner.execute_model(scheduler_output,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/torch/utils/_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_model_runner.py", line 1622, in execute_model
spec_token_ids = self.propose_draft_token_ids(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/dist-packages/vllm/v1/worker/gpu_model_runner.py", line 1745, in propose_draft_token_ids
spec_token_ids = draft_token_ids.tolist()
^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: CUDA error: device-side assert triggered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
terminate called after throwing an instance of 'c10::Error'
ERROR 07-30 20:29:47 [async_llm.py:416] AsyncLLM output_handler failed.
ERROR 07-30 20:29:47 [async_llm.py:416] Traceback (most recent call last):
ERROR 07-30 20:29:47 [async_llm.py:416] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/async_llm.py", line 375, in output_handler
ERROR 07-30 20:29:47 [async_llm.py:416] outputs = await engine_core.get_output_async()
ERROR 07-30 20:29:47 [async_llm.py:416] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [async_llm.py:416] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core_client.py", line 751, in get_output_async
ERROR 07-30 20:29:47 [async_llm.py:416] raise self._format_exception(outputs) from None
ERROR 07-30 20:29:47 [async_llm.py:416] vllm.v1.engine.exceptions.EngineDeadError: EngineCore encountered an issue. See stack trace (above) for the root cause.
INFO 07-30 20:29:47 [async_llm.py:342] Request chatcmpl-698a2381c5aa43e290b3d7bd358817a1 failed (engine dead).
what(): CUDA error: device-side assert triggered
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.
Exception raised from c10_cuda_check_implementation at /pytorch/c10/cuda/CUDAException.cpp:43 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits, std::allocator >) + 0x98 (0x7fe5eb3785e8 in /usr/local/lib/python3.12/dist-packages/torch/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&) + 0xe0 (0x7fe5eb30d4a2 in /usr/local/lib/python3.12/dist-packages/torch/lib/libc10.so)
frame #2: c10::cuda::c10_cuda_check_implementation(int, char const*, char const*, int, bool) + 0x3c2 (0x7fe5eb7ba422 in /usr/local/lib/python3.12/dist-packages/torch/lib/libc10_cuda.so)
frame #3: + 0x1032c05 (0x7fe595c48c05 in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_cuda.so)
frame #4: + 0x102ee8b (0x7fe595c44e8b in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_cuda.so)
frame #5: + 0x10365a4 (0x7fe595c4c5a4 in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_cuda.so)
frame #6: + 0x44d2a2 (0x7fe5e2dc82a2 in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_python.so)
frame #7: c10::TensorImpl::~TensorImpl() + 0x9 (0x7fe5eb352f39 in /usr/local/lib/python3.12/dist-packages/torch/lib/libc10.so)
frame #8: + 0x704f48 (0x7fe5e307ff48 in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_python.so)
frame #9: + 0x705370 (0x7fe5e3080370 in /usr/local/lib/python3.12/dist-packages/torch/lib/libtorch_python.so)
frame #10: /usr/bin/python3.12() [0x59be14]
frame #11: /usr/bin/python3.12() [0x57ccc0]
frame #12: /usr/bin/python3.12() [0x57bcf6]
frame #13: /usr/bin/python3.12() [0x57bcef]
frame #14: /usr/bin/python3.12() [0x57bcef]
frame #15: /usr/bin/python3.12() [0x57bcef]
frame #16: /usr/bin/python3.12() [0x57bcef]
frame #17: /usr/bin/python3.12() [0x57bcef]
frame #18: /usr/bin/python3.12() [0x57bcef]
frame #19: /usr/bin/python3.12() [0x57bcef]
frame #20: /usr/bin/python3.12() [0x57bcef]
frame #21: /usr/bin/python3.12() [0x57bcef]
frame #22: /usr/bin/python3.12() [0x57bcef]
frame #23: /usr/bin/python3.12() [0x57bcef]
frame #24: /usr/bin/python3.12() [0x57bcef]
frame #25: /usr/bin/python3.12() [0x57bcef]
frame #26: /usr/bin/python3.12() [0x57bcef]
frame #27: /usr/bin/python3.12() [0x57bcef]
frame #28: /usr/bin/python3.12() [0x57bcef]
frame #29: /usr/bin/python3.12() [0x594cd7]
frame #30: _PyEval_EvalFrameDefault + 0x50e7 (0x54cb27 in /usr/bin/python3.12)
frame #31: PyEval_EvalCode + 0x99 (0x61d5b9 in /usr/bin/python3.12)
frame #32: /usr/bin/python3.12() [0x6591db]
frame #33: /usr/bin/python3.12() [0x654346]
frame #34: PyRun_StringFlags + 0x63 (0x6503b3 in /usr/bin/python3.12)
frame #35: PyRun_SimpleStringFlags + 0x3e (0x6500be in /usr/bin/python3.12)
frame #36: Py_RunMain + 0x4b2 (0x64d622 in /usr/bin/python3.12)
frame #37: Py_BytesMain + 0x2d (0x6064ad in /usr/bin/python3.12)
frame #38: + 0x29d90 (0x7fe5ec02dd90 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #39: __libc_start_main + 0x80 (0x7fe5ec02de40 in /usr/lib/x86_64-linux-gnu/libc.so.6)
frame #40: _start + 0x25 (0x606325 in /usr/bin/python3.12)
ERROR 07-30 20:29:47 [serving_chat.py:932] Error in chat completion stream generator.
ERROR 07-30 20:29:47 [serving_chat.py:932] Traceback (most recent call last):
ERROR 07-30 20:29:47 [serving_chat.py:932] File "/usr/local/lib/python3.12/dist-packages/vllm/entrypoints/openai/serving_chat.py", line 497, in chat_completion_stream_generator
ERROR 07-30 20:29:47 [serving_chat.py:932] async for res in result_generator:
ERROR 07-30 20:29:47 [serving_chat.py:932] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/async_llm.py", line 323, in generate
ERROR 07-30 20:29:47 [serving_chat.py:932] out = q.get_nowait() or await q.get()
ERROR 07-30 20:29:47 [serving_chat.py:932] ^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [serving_chat.py:932] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/output_processor.py", line 57, in get
ERROR 07-30 20:29:47 [serving_chat.py:932] raise output
ERROR 07-30 20:29:47 [serving_chat.py:932] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/async_llm.py", line 375, in output_handler
ERROR 07-30 20:29:47 [serving_chat.py:932] outputs = await engine_core.get_output_async()
ERROR 07-30 20:29:47 [serving_chat.py:932] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ERROR 07-30 20:29:47 [serving_chat.py:932] File "/usr/local/lib/python3.12/dist-packages/vllm/v1/engine/core_client.py", line 751, in get_output_async
ERROR 07-30 20:29:47 [serving_chat.py:932] raise self._format_exception(outputs) from None
ERROR 07-30 20:29:47 [serving_chat.py:932] vllm.v1.engine.exceptions.EngineDeadError: EngineCore encountered an issue. See stack trace (above) for the root cause.
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: 192.168.0.103:49304 - "POST /v1/chat/completions HTTP/1.1" 500 Internal Server Error
INFO: Shutting down
INFO: Waiting for application shutdown.
INFO: Application shutdown complete.
INFO: Finished server process [73]
Before submitting a new issue...
- Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.