Open osilverstein opened 3 weeks ago
Does this only happen with speculative decoding enabled?
This issue can be reproduced by calling:
The model will erroneously output stop tokens. This does not put other llamas in mode collapse.
It is linked to context length possibly as it does not occur with fewer.
Does this only happen with speculative decoding enabled?
Nope! It happens with long ctx regardless of spec. verified and tested
Update: Tested, this issue does not happen on older version that still supports machete v6.2
Can confirm that my image issue (see comment above) not only resolves with v0.6.1, but also resolves this issue.
Cannot test 0.6.2 due to https://github.com/vllm-project/vllm/issues/8281#issuecomment-2401165333
Can you try with --enforce-eager in 0.6.3? If that works, it's probably related to an issue in 0.6.3 with graph capture + long context that were fixed in https://github.com/vllm-project/vllm/pull/9549
@osilverstein I fixed my issue by installing the latest dev version https://github.com/vllm-project/vllm/issues/9732#issuecomment-2444769412
Gibberish is not produced on the previous version with the same request.
Your current environment
The output of `python collect_env.py`
```plaintext Collecting environment information... 2024-10-28 19:34:18.267757: I tensorflow/core/util/port.cc:153] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. 2024-10-28 19:34:18.467148: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:485] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered 2024-10-28 19:34:18.540226: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:8454] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered 2024-10-28 19:34:18.561701: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1452] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered 2024-10-28 19:34:18.707363: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. To enable the following instructions: AVX2 AVX512F AVX512_VNNI AVX512_BF16 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. 2024-10-28 19:34:19.624862: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT /usr/lib/python3/dist-packages/scipy/__init__.py:146: UserWarning: A NumPy version >=1.17.3 and <1.25.0 is required for this version of SciPy (detected version 1.26.4 warnings.warn(f"A NumPy version >={np_minversion} and <{np_maxversion}" 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: 3.10.12 (main, Sep 11 2024, 15:47:36) [GCC 11.4.0] (64-bit runtime) Python platform: Linux-5.15.0-122-generic-x86_64-with-glibc2.35 Is CUDA available: True CUDA runtime version: 12.4.131 CUDA_MODULE_LOADING set to: LAZY GPU models and configuration: GPU 0: NVIDIA H200 GPU 1: NVIDIA H200 GPU 2: NVIDIA H200 GPU 3: NVIDIA H200 GPU 4: NVIDIA H200 GPU 5: NVIDIA H200 GPU 6: NVIDIA H200 GPU 7: NVIDIA H200 Nvidia driver version: 550.90.12 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_adv.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_cnn.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_engines_precompiled.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_engines_runtime_compiled.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_graph.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_heuristic.so.9.5.0 /usr/lib/x86_64-linux-gnu/libcudnn_ops.so.9.5.0 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: 52 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 384 On-line CPU(s) list: 0-383 Vendor ID: AuthenticAMD Model name: AMD EPYC 9654 96-Core Processor CPU family: 25 Model: 17 Thread(s) per core: 2 Core(s) per socket: 96 Socket(s): 2 Stepping: 1 Frequency boost: enabled CPU max MHz: 3707.8120 CPU min MHz: 1500.0000 BogoMIPS: 4800.15 Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 invpcid_single hw_pstate ssbd mba ibrs ibpb stibp ibrs_enhanced vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local avx512_bf16 clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq la57 rdpid overflow_recov succor smca fsrm flush_l1d Virtualization: AMD-V L1d cache: 6 MiB (192 instances) L1i cache: 6 MiB (192 instances) L2 cache: 192 MiB (192 instances) L3 cache: 768 MiB (24 instances) NUMA node(s): 2 NUMA node0 CPU(s): 0-95,192-287 NUMA node1 CPU(s): 96-191,288-383 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: Not affected Vulnerability Retbleed: Not affected Vulnerability Spec rstack overflow: Mitigation; safe RET 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 / Automatic IBRS; IBPB conditional; STIBP always-on; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected 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==12.4.99 [pip3] nvidia-nvtx-cu12==12.1.105 [pip3] optree==0.12.1 [pip3] pyzmq==26.2.0 [pip3] torch==2.4.0 [pip3] torchaudio==2.4.1+cu124 [pip3] torchvision==0.19.0 [pip3] transformers==4.46.0 [pip3] triton==3.0.0 [conda] Could not collect ROCM Version: Could not collect Neuron SDK Version: N/A vLLM Version: 0.6.3.post1 vLLM Build Flags: CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled GPU Topology: GPU0 GPU1 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 CPU Affinity NUMA Affinity GPU NUMA ID GPU0 X NV18 NV18 NV18 NV18 NV18 NV18 NV18 SYS SYS 0-95,192-287 0 N/A GPU1 NV18 X NV18 NV18 NV18 NV18 NV18 NV18 SYS SYS 0-95,192-287 0 N/A GPU2 NV18 NV18 X NV18 NV18 NV18 NV18 NV18 SYS SYS 0-95,192-287 0 N/A GPU3 NV18 NV18 NV18 X NV18 NV18 NV18 NV18 SYS SYS 0-95,192-287 0 N/A GPU4 NV18 NV18 NV18 NV18 X NV18 NV18 NV18 NODE NODE 96-191,288-383 1 N/A GPU5 NV18 NV18 NV18 NV18 NV18 X NV18 NV18 PIX PIX 96-191,288-383 1 N/A GPU6 NV18 NV18 NV18 NV18 NV18 NV18 X NV18 NODE NODE 96-191,288-383 1 N/A GPU7 NV18 NV18 NV18 NV18 NV18 NV18 NV18 X NODE NODE 96-191,288-383 1 N/A NIC0 SYS SYS SYS SYS NODE PIX NODE NODE X PIX NIC1 SYS SYS SYS SYS NODE PIX NODE NODE 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 ```🐛 Describe the bug
I start the inference server with:
Note the use of speculative decoding, tensor parallelism, and the draft parallel size.
When running our pipeline, we occasionally get outputs that are clearly just random tokens strewn together. I can provide an example, but essentially, ~1 in 10 requests return gibberish regardless of content. (It gets worse the longer it is, I think). There are no error messages or warnings.
I was wondering if anyone had encountered this issue or if someone at Neural Magic has any intuition for what could be causing it. Perhaps this has been addressed with the pending addition of parallel drafting?
@alexm-neuralmagic @robertgshaw2-neuralmagic
Before submitting a new issue...