[Bug]: I am not able to start the vllm container with llama 3.1 70b

[pranjalst]

Your current environment

/usr/lib/python3.10/inspect.py:288: FutureWarning: torch.distributed.reduce_op is deprecated, please use torch.distributed.ReduceOp instead
return isinstance(object, types.FunctionType)
Collecting environment information...
PyTorch version: 2.4.0a0+git74cd574
Is debug build: False
CUDA used to build PyTorch: None
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 3.22.1
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: False
CUDA runtime version: No CUDA
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration: No CUDA
Nvidia driver version: No CUDA
cuDNN version: No CUDA
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): 152
On-line CPU(s) list: 0-151
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8368 CPU @ 2.40GHz
CPU family: 6
Model: 106
Thread(s) per core: 2
Core(s) per socket: 38
Socket(s): 2
Stepping: 6
CPU max MHz: 3400.0000
CPU min MHz: 800.0000
BogoMIPS: 4800.00
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 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 invpcid_single 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 wbnoinvd dtherm ida arat pln pts avx512vbmi umip pku ospke avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg tme avx512_vpopcntdq la57 rdpid fsrm md_clear pconfig flush_l1d arch_capabilities
Virtualization: VT-x
L1d cache: 3.6 MiB (76 instances)
L1i cache: 2.4 MiB (76 instances)
L2 cache: 95 MiB (76 instances)
L3 cache: 114 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0-37,76-113
NUMA node1 CPU(s): 38-75,114-151
Vulnerability Gather data sampling: Mitigation; Microcode
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Mmio stale data: Mitigation; Clear CPU buffers; SMT vulnerable
Vulnerability Reg file data sampling: 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 / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI SW loop, KVM SW loop
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected

Versions of relevant libraries:
[pip3] habana-torch-dataloader==1.18.0.524
[pip3] habana-torch-plugin==1.18.0.524
[pip3] numpy==1.26.4
[pip3] pynvml==8.0.4
[pip3] pytorch-lightning==2.4.0
[pip3] pyzmq==26.2.0
[pip3] torch==2.4.0a0+git74cd574
[pip3] torch_tb_profiler==0.4.0
[pip3] torchaudio==2.4.0a0+69d4077
[pip3] torchdata==0.7.1+5e6f7b7
[pip3] torchmetrics==1.4.2
[pip3] torchtext==0.18.0a0+9bed85d
[pip3] torchvision==0.19.0a0+48b1edf
[pip3] transformers==4.46.2
[pip3] triton==3.1.0
[conda] Could not collect
ROCM Version: Could not collect
Neuron SDK Version: N/A
vLLM Version: 0.6.3.dev915+g41dddabb.d20241111
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
Could not collect

Model Input Dumps

No response

🐛 Describe the bug

I'm facing an issue when running the vllm OpenAI API server with the following command:
/bin/bash -c "export VLLM_CPU_KVCACHE_SPACE=40 && python3 -m vllm.entrypoints.openai.api_server --enforce-eager --model meta-llama/Llama-3.1-70B-Instruct --tensor-parallel-size 8 --max-model-len 32000 --max-num-batched-tokens 128000 --host 0.0.0.0 --port 80" and i am facing issue

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.

[michalkuligowski]

@pranjalst please use vllm-fork branch v0.5.3.post1+Gaudi-1.18.0 for 1.18 release as in https://github.com/HabanaAI/vllm-fork/releases/tag/v0.5.3.post1%2BGaudi-1.18.0

[pranjalst]

i am using that one..

[michalkuligowski]

@pranjalst from your collect_env output - "vLLM Version: 0.6.3.dev915+g41dddabb.d20241111" and you should have something like
"vLLM Version: 0.5.3.post1"

[pranjalst]

i am using 0.5.3.post 1 still i am facing the issue

[iboiko-habana]

Please set next flags for OOM/functional issues avoiding in 1.18.0
VLLM_ENGINE_ITERATION_TIMEOUT_S=3600
VLLM_RPC_TIMEOUT=100000
VLLM_PROMPT_USE_FUSEDSDPA=1
PT_HPU_ENABLE_LAZY_COLLECTIVES=true

Other flags, depending on context length. 32K context length flags example:

decreasing of VLLM_GRAPH_RESERVED_MEM, depends on model and long context. VLLM_GRAPH_RESERVED_MEM=0.02 for llama3.1-8b. VLLM_GRAPH_RESERVED_MEM=0.1 for llama3.1-70b.
VLLM_PROMPT_BS_BUCKET_MIN=1 # proposal for usage. depends on model. Can be increased if no OOM
VLLM_PROMPT_BS_BUCKET_STEP=16 # proposal for usage. depends on model. Can be increased until no OOM or decreased if OOM
VLLM_PROMPT_BS_BUCKET_MAX=16 # proposal for usage. depends on model. Can be increased until no OOM or decreased if OOM
VLLM_PROMPT_SEQ_BUCKET_MIN=24576 # proposal for usage. depends on warmup results
VLLM_PROMPT_SEQ_BUCKET_STEP=2048 # proposal for usage. depends on warmup results
VLLM_PROMPT_SEQ_BUCKET_MAX=32768 # context length 32K, 16384 for 16K
VLLM_DECODE_BLOCK_BUCKET_MIN=1024 # proposal for usage. depends on warmup results
VLLM_DECODE_BLOCK_BUCKET_STEP=1024 # proposal for usage. depends on warmup results
VLLM_DECODE_BLOCK_BUCKET_MAX=max_num_seqs * max_decode_seq // self.block_size # i.e. 128*(32 * 1024)/128 or 32*(32*1024)/128

[michalkuligowski]

hi @pranjalst did @iboiko-habana comment help?