vllm: [Bug]: guided_json bad output for llama2-13b

Your current environment

Collecting environment information...
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
ROCM used to build PyTorch: N/A
 
OS: Ubuntu 22.04.3 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: 3.10.9 (main, Jan 11 2023, 15:21:40) [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-5.15.0-89-generic-x86_64-with-glibc2.35
Is CUDA available: N/A
CUDA runtime version: 12.1.66
CUDA_MODULE_LOADING set to: N/A
GPU models and configuration:
GPU 0: NVIDIA A100-SXM4-80GB
GPU 1: NVIDIA A100-SXM4-80GB
GPU 2: NVIDIA A100-SXM4-80GB
GPU 3: NVIDIA A100-SXM4-80GB
GPU 4: NVIDIA A100-SXM4-80GB
GPU 5: NVIDIA A100-SXM4-80GB
GPU 6: NVIDIA A100-SXM4-80GB
GPU 7: NVIDIA A100-SXM4-80GB
 
Nvidia driver version: 550.54.14
cuDNN version: Could not collect
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: N/A
 
CPU:
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): 192
On-line CPU(s) list: 0-191
Vendor ID: GenuineIntel
Model name: Intel(R) Xeon(R) Platinum 8468
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 48
Socket(s): 2
Stepping: 8
BogoMIPS: 4200.00
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 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 split_lock_detect 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: 4.5 MiB (96 instances)
L1i cache: 3 MiB (96 instances)
L2 cache: 192 MiB (96 instances)
L3 cache: 210 MiB (2 instances)
NUMA node(s): 2
NUMA node0 CPU(s): 0,2,4,6,8,10,12,14,16,18,20,22,24,26,28,30,32,34,36,38,40,42,44,46,48,50,52,54,56,58,60,62,64,66,68,70,72,74,76,78,80,82,84,86,88,90,92,94,96,98,100,102,104,106,108,110,112,114,116,118,120,122,124,126,128,130,132,134,136,138,140,142,144,146,148,150,152,154,156,158,160,162,164,166,168,170,172,174,176,178,180,182,184,186,188,190
NUMA node1 CPU(s): 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31,33,35,37,39,41,43,45,47,49,51,53,55,57,59,61,63,65,67,69,71,73,75,77,79,81,83,85,87,89,91,93,95,97,99,101,103,105,107,109,111,113,115,117,119,121,123,125,127,129,131,133,135,137,139,141,143,145,147,149,151,153,155,157,159,161,163,165,167,169,171,173,175,177,179,181,183,185,187,189,191
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 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 IBRS, IBPB conditional, RSB filling, PBRSB-eIBRS SW sequence
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
 
Versions of relevant libraries:
[pip3] No relevant packages
[conda] No relevant packagesROCM 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 GPU2 GPU3 GPU4 GPU5 GPU6 GPU7 NIC0 NIC1 NIC2 NIC3 NIC4 NIC5 NIC6 CPU Affinity NUMA Affinity GPU NUMA ID
GPU0 X NV12 NV12 NV12 NV12 NV12 NV12 NV12 PXB PXB PXB SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A
GPU1 NV12 X NV12 NV12 NV12 NV12 NV12 NV12 SYS SYS SYS SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A
GPU2 NV12 NV12 X NV12 NV12 NV12 NV12 NV12 SYS SYS SYS SYS SYS SYS SYS 0,2,4,6,8,10 0 N/A
GPU3 NV12 NV12 NV12 X NV12 NV12 NV12 NV12 SYS SYS SYS PXB SYS SYS SYS 0,2,4,6,8,10 0 N/A
GPU4 NV12 NV12 NV12 NV12 X NV12 NV12 NV12 SYS SYS SYS SYS PXB SYS PXB 1,3,5,7,9,11 1 N/A
GPU5 NV12 NV12 NV12 NV12 NV12 X NV12 NV12 SYS SYS SYS SYS SYS SYS SYS 1,3,5,7,9,11 1 N/A
GPU6 NV12 NV12 NV12 NV12 NV12 NV12 X NV12 SYS SYS SYS SYS SYS PXB SYS 1,3,5,7,9,11 1 N/A
GPU7 NV12 NV12 NV12 NV12 NV12 NV12 NV12 X SYS SYS SYS SYS SYS SYS SYS 1,3,5,7,9,11 1 N/A
NIC0 PXB SYS SYS SYS SYS SYS SYS SYS X PIX PIX SYS SYS SYS SYS
NIC1 PXB SYS SYS SYS SYS SYS SYS SYS PIX X PIX SYS SYS SYS SYS
NIC2 PXB SYS SYS SYS SYS SYS SYS SYS PIX PIX X SYS SYS SYS SYS
NIC3 SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS SYS SYS
NIC4 SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS X SYS PIX
NIC5 SYS SYS SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS X SYS
NIC6 SYS SYS SYS SYS PXB SYS SYS SYS SYS SYS SYS SYS PIX SYS 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
  NIC2: mlx5_2
  NIC3: mlx5_3
  NIC4: mlx5_4
  NIC5: mlx5_7
  NIC6: mlx5_bond_0
 

🐛 Describe the bug

Using vllm 0.4.0.post1

import openai
 
client = openai.OpenAI(base_url='')
 
kwargs = {'model': 'h2oai/h2ogpt-4096-llama2-13b-chat', 'prompt': '[INST] \n\nGive an example employee profile. [/INST]', 'temperature': 0, 'max_tokens': 1024, 'top_p': 1, 'frequency_penalty': 0, 'seed': 1, 'n': 1, 'presence_penalty': 0.14000000000000012, 'stop': ['[/INST]', '', '[INST]'], 'extra_body': {'stop_token_ids': [2, 2], 'response_format': {'type': 'json_object'}, 'guided_json': {'type': 'object', 'properties': {'name': {'type': 'string'}, 'age': {'type': 'integer'}, 'skills': {'type': 'array', 'items': {'type': 'string', 'maxLength': 10}, 'minItems': 3}, 'workhistory': {'type': 'array', 'items': {'type': 'object', 'properties': {'company': {'type': 'string'}, 'duration': {'type': 'string'}, 'position': {'type': 'string'}}, 'required': ['company', 'position']}}}, 'required': ['name', 'age', 'skills', 'workhistory']}}, 'stream': False, 'timeout': 600}
response = client.completions.create(**kwargs).choices[0].text
print(response)

streaming or not doesn’t matter, happens every time same way. Get this response:

{
 
"age":
 
   
 
   
....

Here … means a MASSIVE number of new lines.

I’ve tried about 10 other models, all are ok except llama2-13b.

Others that work:

['h2oai/h2ogpt-4096-llama2-70b-chat',
'mistralai/Mixtral-8x7B-Instruct-v0.1',
                     'HuggingFaceH4/zephyr-7b-beta', 'mistralai/Mistral-7B-Instruct-v0.2', 'openchat/openchat-3.5-1210',
                     'h2oai/h2ogpt-32k-codellama-34b-instruct', 'NousResearch/Nous-Capybara-34B',
                     'h2oai/h2o-danube2-1.8b-chat',
                     'google/gemma-1.1-7b-it',
                     'databricks/dbrx-instruct', 'CohereForAI/c4ai-command-r-v01',]

About this issue

  • Original URL
  • State: closed
  • Created 3 years ago
  • Comments: 20 (2 by maintainers)

Commits related to this issue

Most upvoted comments

@rlouf FYI I h it this even with ‘mistralai/Mixtral-8x7B-Instruct-v0.1’ even when providing the schema to the model. Putting ‘[ \t\n]’ failed and ‘\n’ failed to work, but ’ ’ did ok like @robcaulk mentioned above.

I ran into this problem with Llama3 8b instruct and solved it by running this PR #4305 and setting guided_whitespace_pattern to " ".

Thank you for the report. I think this is because newline is allowed and the model can indefinitely spam new lines if it choose to be. This commonly occurs when the model is not well prompted or finetuned for this workload. Even with guided JSON, give a few shot examples could be helpful. So prompting would be the simplest fix.

I can think of two slightly more complex fix:

  • Some where in outlines (cc @rlouf) whitespace repeated is allowed between JSON key and values, maybe we should just set it to one?
  • Maybe we can consider repetition penalty, but this will change output accuracy as it will also apply things like values.